Search Results

Search found 6178 results on 248 pages for 'section'.

Page 48/248 | < Previous Page | 44 45 46 47 48 49 50 51 52 53 54 55  | Next Page >

• Currency Conversion in Oracle BI applications

  - by Saurabh Verma

  Authored by Vijay Aggarwal and Hichem Sellami A typical data warehouse contains Star and/or Snowflake schema, made up of Dimensions and Facts. The facts store various numerical information including amounts. Example; Order Amount, Invoice Amount etc. With the true global nature of business now-a-days, the end-users want to view the reports in their own currency or in global/common currency as defined by their business. This presents a unique opportunity in BI to provide the amounts in converted rates either by pre-storing or by doing on-the-fly conversions while displaying the reports to the users. Source Systems OBIA caters to various source systems like EBS, PSFT, Sebl, JDE, Fusion etc. Each source has its own unique and intricate ways of defining and storing currency data, doing currency conversions and presenting to the OLTP users. For example; EBS stores conversion rates between currencies which can be classified by conversion rates, like Corporate rate, Spot rate, Period rate etc. Siebel stores exchange rates by conversion rates like Daily. EBS/Fusion stores the conversion rates for each day, where as PSFT/Siebel store for a range of days. PSFT has Rate Multiplication Factor and Rate Division Factor and we need to calculate the Rate based on them, where as other Source systems store the Currency Exchange Rate directly. OBIA Design The data consolidation from various disparate source systems, poses the challenge to conform various currencies, rate types, exchange rates etc., and designing the best way to present the amounts to the users without affecting the performance. When consolidating the data for reporting in OBIA, we have designed the mechanisms in the Common Dimension, to allow users to report based on their required currencies. OBIA Facts store amounts in various currencies: Document Currency: This is the currency of the actual transaction. For a multinational company, this can be in various currencies. Local Currency: This is the base currency in which the accounting entries are recorded by the business. This is generally defined in the Ledger of the company. Global Currencies: OBIA provides five Global Currencies. Three are used across all modules. The last two are for CRM only. A Global currency is very useful when creating reports where the data is viewed enterprise-wide. Example; a US based multinational would want to see the reports in USD. The company will choose USD as one of the global currencies. OBIA allows users to define up-to five global currencies during the initial implementation. The term Currency Preference is used to designate the set of values: Document Currency, Local Currency, Global Currency 1, Global Currency 2, Global Currency 3; which are shared among all modules. There are four more currency preferences, specific to certain modules: Global Currency 4 (aka CRM Currency) and Global Currency 5 which are used in CRM; and Project Currency and Contract Currency, used in Project Analytics. When choosing Local Currency for Currency preference, the data will show in the currency of the Ledger (or Business Unit) in the prompt. So it is important to select one Ledger or Business Unit when viewing data in Local Currency. More on this can be found in the section: Toggling Currency Preferences in the Dashboard. Design Logic When extracting the fact data, the OOTB mappings extract and load the document amount, and the local amount in target tables. It also loads the exchange rates required to convert the document amount into the corresponding global amounts. If the source system only provides the document amount in the transaction, the extract mapping does a lookup to get the Local currency code, and the Local exchange rate. The Load mapping then uses the local currency code and rate to derive the local amount. The load mapping also fetches the Global Currencies and looks up the corresponding exchange rates. The lookup of exchange rates is done via the Exchange Rate Dimension provided as a Common/Conforming Dimension in OBIA. The Exchange Rate Dimension stores the exchange rates between various currencies for a date range and Rate Type. Two physical tables W_EXCH_RATE_G and W_GLOBAL_EXCH_RATE_G are used to provide the lookups and conversions between currencies. The data is loaded from the source system’s Ledger tables. W_EXCH_RATE_G stores the exchange rates between currencies with a date range. On the other hand, W_GLOBAL_EXCH_RATE_G stores the currency conversions between the document currency and the pre-defined five Global Currencies for each day. Based on the requirements, the fact mappings can decide and use one or both tables to do the conversion. Currency design in OBIA also taps into the MLS and Domain architecture, thus allowing the users to map the currencies to a universal Domain during the implementation time. This is especially important for companies deploying and using OBIA with multiple source adapters. Some Gotchas to Look for It is necessary to think through the currencies during the initial implementation. 1) Identify various types of currencies that are used by your business. Understand what will be your Local (or Base) and Documentation currency. Identify various global currencies that your users will want to look at the reports. This will be based on the global nature of your business. Changes to these currencies later in the project, while permitted, but may cause Full data loads and hence lost time. 2) If the user has a multi source system make sure that the Global Currencies and Global Rate Types chosen in Configuration Manager do have the corresponding source specific counterparts. In other words, make sure for every DW specific value chosen for Currency Code or Rate Type, there is a source Domain mapping already done. Technical Section This section will briefly mention the technical scenarios employed in the OBIA adaptors to extract data from each source system. In OBIA, we have two main tables which store the Currency Rate information as explained in previous sections. W_EXCH_RATE_G and W_GLOBAL_EXCH_RATE_G are the two tables. W_EXCH_RATE_G stores all the Currency Conversions present in the source system. It captures data for a Date Range. W_GLOBAL_EXCH_RATE_G has Global Currency Conversions stored at a Daily level. However the challenge here is to store all the 5 Global Currency Exchange Rates in a single record for each From Currency. Let’s voyage further into the Source System Extraction logic for each of these tables and understand the flow briefly. EBS: In EBS, we have Currency Data stored in GL_DAILY_RATES table. As the name indicates GL_DAILY_RATES EBS table has data at a daily level. However in our warehouse we store the data with a Date Range and insert a new range record only when the Exchange Rate changes for a particular From Currency, To Currency and Rate Type. Below are the main logical steps that we employ in this process. (Incremental Flow only) – Cleanup the data in W_EXCH_RATE_G. Delete the records which have Start Date > minimum conversion date Update the End Date of the existing records. Compress the daily data from GL_DAILY_RATES table into Range Records. Incremental map uses $$XRATE_UPD_NUM_DAY as an extra parameter. Generate Previous Rate, Previous Date and Next Date for each of the Daily record from the OLTP. Filter out the records which have Conversion Rate same as Previous Rates or if the Conversion Date lies within a single day range. Mark the records as ‘Keep’ and ‘Filter’ and also get the final End Date for the single Range record (Unique Combination of From Date, To Date, Rate and Conversion Date). Filter the records marked as ‘Filter’ in the INFA map. The above steps will load W_EXCH_RATE_GS. Step 0 updates/deletes W_EXCH_RATE_G directly. SIL map will then insert/update the GS data into W_EXCH_RATE_G. These steps convert the daily records in GL_DAILY_RATES to Range records in W_EXCH_RATE_G. We do not need such special logic for loading W_GLOBAL_EXCH_RATE_G. This is a table where we store data at a Daily Granular Level. However we need to pivot the data because the data present in multiple rows in source tables needs to be stored in different columns of the same row in DW. We use GROUP BY and CASE logic to achieve this. Fusion: Fusion has extraction logic very similar to EBS. The only difference is that the Cleanup logic that was mentioned in step 0 above does not use $$XRATE_UPD_NUM_DAY parameter. In Fusion we bring all the Exchange Rates in Incremental as well and do the cleanup. The SIL then takes care of Insert/Updates accordingly. PeopleSoft:PeopleSoft does not have From Date and To Date explicitly in the Source tables. Let’s look at an example. Please note that this is achieved from PS1 onwards only. 1 Jan 2010 – USD to INR – 45 31 Jan 2010 – USD to INR – 46 PSFT stores records in above fashion. This means that Exchange Rate of 45 for USD to INR is applicable for 1 Jan 2010 to 30 Jan 2010. We need to store data in this fashion in DW. Also PSFT has Exchange Rate stored as RATE_MULT and RATE_DIV. We need to do a RATE_MULT/RATE_DIV to get the correct Exchange Rate. We generate From Date and To Date while extracting data from source and this has certain assumptions: If a record gets updated/inserted in the source, it will be extracted in incremental. Also if this updated/inserted record is between other dates, then we also extract the preceding and succeeding records (based on dates) of this record. This is required because we need to generate a range record and we have 3 records whose ranges have changed. Taking the same example as above, if there is a new record which gets inserted on 15 Jan 2010; the new ranges are 1 Jan to 14 Jan, 15 Jan to 30 Jan and 31 Jan to Next available date. Even though 1 Jan record and 31 Jan have not changed, we will still extract them because the range is affected. Similar logic is used for Global Exchange Rate Extraction. We create the Range records and get it into a Temporary table. Then we join to Day Dimension, create individual records and pivot the data to get the 5 Global Exchange Rates for each From Currency, Date and Rate Type. Siebel: Siebel Facts are dependent on Global Exchange Rates heavily and almost none of them really use individual Exchange Rates. In other words, W_GLOBAL_EXCH_RATE_G is the main table used in Siebel from PS1 release onwards. As of January 2002, the Euro Triangulation method for converting between currencies belonging to EMU members is not needed for present and future currency exchanges. However, the method is still available in Siebel applications, as are the old currencies, so that historical data can be maintained accurately. The following description applies only to historical data needing conversion prior to the 2002 switch to the Euro for the EMU member countries. If a country is a member of the European Monetary Union (EMU), you should convert its currency to other currencies through the Euro. This is called triangulation, and it is used whenever either currency being converted has EMU Triangulation checked. Due to this, there are multiple extraction flows in SEBL ie. EUR to EMU, EUR to NonEMU, EUR to DMC and so on. We load W_EXCH_RATE_G through multiple flows with these data. This has been kept same as previous versions of OBIA. W_GLOBAL_EXCH_RATE_G being a new table does not have such needs. However SEBL does not have From Date and To Date columns in the Source tables similar to PSFT. We use similar extraction logic as explained in PSFT section for SEBL as well. What if all 5 Global Currencies configured are same? As mentioned in previous sections, from PS1 onwards we store Global Exchange Rates in W_GLOBAL_EXCH_RATE_G table. The extraction logic for this table involves Pivoting data from multiple rows into a single row with 5 Global Exchange Rates in 5 columns. As mentioned in previous sections, we use CASE and GROUP BY functions to achieve this. This approach poses a unique problem when all the 5 Global Currencies Chosen are same. For example – If the user configures all 5 Global Currencies as ‘USD’ then the extract logic will not be able to generate a record for From Currency=USD. This is because, not all Source Systems will have a USD->USD conversion record. We have _Generated mappings to take care of this case. We generate a record with Conversion Rate=1 for such cases. Reusable Lookups Before PS1, we had a Mapplet for Currency Conversions. In PS1, we only have reusable Lookups- LKP_W_EXCH_RATE_G and LKP_W_GLOBAL_EXCH_RATE_G. These lookups have another layer of logic so that all the lookup conditions are met when they are used in various Fact Mappings. Any user who would want to do a LKP on W_EXCH_RATE_G or W_GLOBAL_EXCH_RATE_G should and must use these Lookups. A direct join or Lookup on the tables might lead to wrong data being returned. Changing Currency preferences in the Dashboard: In the 796x series, all amount metrics in OBIA were showing the Global1 amount. The customer needed to change the metric definitions to show them in another Currency preference. Project Analytics started supporting currency preferences since 7.9.6 release though, and it published a Tech note for other module customers to add toggling between currency preferences to the solution. List of Currency Preferences Starting from 11.1.1.x release, the BI Platform added a new feature to support multiple currencies. The new session variable (PREFERRED_CURRENCY) is populated through a newly introduced currency prompt. This prompt can take its values from the xml file: userpref_currencies_OBIA.xml, which is hosted in the BI Server installation folder, under :< home>\instances\instance1\config\OracleBIPresentationServicesComponent\coreapplication_obips1\userpref_currencies.xml This file contains the list of currency preferences, like“Local Currency”, “Global Currency 1”,…which customers can also rename to give them more meaningful business names. There are two options for showing the list of currency preferences to the user in the dashboard: Static and Dynamic. In Static mode, all users will see the full list as in the user preference currencies file. In the Dynamic mode, the list shown in the currency prompt drop down is a result of a dynamic query specified in the same file. Customers can build some security into the rpd, so the list of currency preferences will be based on the user roles…BI Applications built a subject area: “Dynamic Currency Preference” to run this query, and give every user only the list of currency preferences required by his application roles. Adding Currency to an Amount Field When the user selects one of the items from the currency prompt, all the amounts in that page will show in the Currency corresponding to that preference. For example, if the user selects “Global Currency1” from the prompt, all data will be showing in Global Currency 1 as specified in the Configuration Manager. If the user select “Local Currency”, all amount fields will show in the Currency of the Business Unit selected in the BU filter of the same page. If there is no particular Business Unit selected in that filter, and the data selected by the query contains amounts in more than one currency (for example one BU has USD as a functional currency, the other has EUR as functional currency), then subtotals will not be available (cannot add USD and EUR amounts in one field), and depending on the set up (see next paragraph), the user may receive an error. There are two ways to add the Currency field to an amount metric: In the form of currency code, like USD, EUR…For this the user needs to add the field “Apps Common Currency Code” to the report. This field is in every subject area, usually under the table “Currency Tag” or “Currency Code”… In the form of currency symbol ($ for USD, € for EUR,…) For this, the user needs to format the amount metrics in the report as a currency column, by specifying the currency tag column in the Column Properties option in Column Actions drop down list. Typically this column should be the “BI Common Currency Code” available in every subject area. Select Column Properties option in the Edit list of a metric. In the Data Format tab, select Custom as Treat Number As. Enter the following syntax under Custom Number Format: [$:currencyTagColumn=Subjectarea.table.column] Where Column is the “BI Common Currency Code” defined to take the currency code value based on the currency preference chosen by the user in the Currency preference prompt.

  Read the article

• How John Got 15x Improvement Without Really Trying

  - by rchrd

  The following article was published on a Sun Microsystems website a number of years ago by John Feo. It is still useful and worth preserving. So I'm republishing it here.  How I Got 15x Improvement Without Really Trying John Feo, Sun Microsystems Taking ten "personal" program codes used in scientific and engineering research, the author was able to get from 2 to 15 times performance improvement easily by applying some simple general optimization techniques. Introduction Scientific research based on computer simulation depends on the simulation for advancement. The research can advance only as fast as the computational codes can execute. The codes' efficiency determines both the rate and quality of results. In the same amount of time, a faster program can generate more results and can carry out a more detailed simulation of physical phenomena than a slower program. Highly optimized programs help science advance quickly and insure that monies supporting scientific research are used as effectively as possible. Scientific computer codes divide into three broad categories: ISV, community, and personal. ISV codes are large, mature production codes developed and sold commercially. The codes improve slowly over time both in methods and capabilities, and they are well tuned for most vendor platforms. Since the codes are mature and complex, there are few opportunities to improve their performance solely through code optimization. Improvements of 10% to 15% are typical. Examples of ISV codes are DYNA3D, Gaussian, and Nastran. Community codes are non-commercial production codes used by a particular research field. Generally, they are developed and distributed by a single academic or research institution with assistance from the community. Most users just run the codes, but some develop new methods and extensions that feed back into the general release. The codes are available on most vendor platforms. Since these codes are younger than ISV codes, there are more opportunities to optimize the source code. Improvements of 50% are not unusual. Examples of community codes are AMBER, CHARM, BLAST, and FASTA. Personal codes are those written by single users or small research groups for their own use. These codes are not distributed, but may be passed from professor-to-student or student-to-student over several years. They form the primordial ocean of applications from which community and ISV codes emerge. Government research grants pay for the development of most personal codes. This paper reports on the nature and performance of this class of codes. Over the last year, I have looked at over two dozen personal codes from more than a dozen research institutions. The codes cover a variety of scientific fields, including astronomy, atmospheric sciences, bioinformatics, biology, chemistry, geology, and physics. The sources range from a few hundred lines to more than ten thousand lines, and are written in Fortran, Fortran 90, C, and C++. For the most part, the codes are modular, documented, and written in a clear, straightforward manner. They do not use complex language features, advanced data structures, programming tricks, or libraries. I had little trouble understanding what the codes did or how data structures were used. Most came with a makefile. Surprisingly, only one of the applications is parallel. All developers have access to parallel machines, so availability is not an issue. Several tried to parallelize their applications, but stopped after encountering difficulties. Lack of education and a perception that parallelism is difficult prevented most from trying. I parallelized several of the codes using OpenMP, and did not judge any of the codes as difficult to parallelize. Even more surprising than the lack of parallelism is the inefficiency of the codes. I was able to get large improvements in performance in a matter of a few days applying simple optimization techniques. Table 1 lists ten representative codes [names and affiliation are omitted to preserve anonymity]. Improvements on one processor range from 2x to 15.5x with a simple average of 4.75x. I did not use sophisticated performance tools or drill deep into the program's execution character as one would do when tuning ISV or community codes. Using only a profiler and source line timers, I identified inefficient sections of code and improved their performance by inspection. The changes were at a high level. I am sure there is another factor of 2 or 3 in each code, and more if the codes are parallelized. The study’s results show that personal scientific codes are running many times slower than they should and that the problem is pervasive. Computational scientists are not sloppy programmers; however, few are trained in the art of computer programming or code optimization. I found that most have a working knowledge of some programming language and standard software engineering practices; but they do not know, or think about, how to make their programs run faster. They simply do not know the standard techniques used to make codes run faster. In fact, they do not even perceive that such techniques exist. The case studies described in this paper show that applying simple, well known techniques can significantly increase the performance of personal codes. It is important that the scientific community and the Government agencies that support scientific research find ways to better educate academic scientific programmers. The inefficiency of their codes is so bad that it is retarding both the quality and progress of scientific research. # cacheperformance redundantoperations loopstructures performanceimprovement 1 x x 15.5 2 x 2.8 3 x x 2.5 4 x 2.1 5 x x 2.0 6 x 5.0 7 x 5.8 8 x 6.3 9 2.2 10 x x 3.3 Table 1 — Area of improvement and performance gains of 10 codes The remainder of the paper is organized as follows: sections 2, 3, and 4 discuss the three most common sources of inefficiencies in the codes studied. These are cache performance, redundant operations, and loop structures. Each section includes several examples. The last section summaries the work and suggests a possible solution to the issues raised. Optimizing cache performance Commodity microprocessor systems use caches to increase memory bandwidth and reduce memory latencies. Typical latencies from processor to L1, L2, local, and remote memory are 3, 10, 50, and 200 cycles, respectively. Moreover, bandwidth falls off dramatically as memory distances increase. Programs that do not use cache effectively run many times slower than programs that do. When optimizing for cache, the biggest performance gains are achieved by accessing data in cache order and reusing data to amortize the overhead of cache misses. Secondary considerations are prefetching, associativity, and replacement; however, the understanding and analysis required to optimize for the latter are probably beyond the capabilities of the non-expert. Much can be gained simply by accessing data in the correct order and maximizing data reuse. 6 out of the 10 codes studied here benefited from such high level optimizations. Array Accesses The most important cache optimization is the most basic: accessing Fortran array elements in column order and C array elements in row order. Four of the ten codes—1, 2, 4, and 10—got it wrong. Compilers will restructure nested loops to optimize cache performance, but may not do so if the loop structure is too complex, or the loop body includes conditionals, complex addressing, or function calls. In code 1, the compiler failed to invert a key loop because of complex addressing do I = 0, 1010, delta_x IM = I - delta_x IP = I + delta_x do J = 5, 995, delta_x JM = J - delta_x JP = J + delta_x T1 = CA1(IP, J) + CA1(I, JP) T2 = CA1(IM, J) + CA1(I, JM) S1 = T1 + T2 - 4 * CA1(I, J) CA(I, J) = CA1(I, J) + D * S1 end do end do In code 2, the culprit is conditionals do I = 1, N do J = 1, N If (IFLAG(I,J) .EQ. 0) then T1 = Value(I, J-1) T2 = Value(I-1, J) T3 = Value(I, J) T4 = Value(I+1, J) T5 = Value(I, J+1) Value(I,J) = 0.25 * (T1 + T2 + T5 + T4) Delta = ABS(T3 - Value(I,J)) If (Delta .GT. MaxDelta) MaxDelta = Delta endif enddo enddo I fixed both programs by inverting the loops by hand. Code 10 has three-dimensional arrays and triply nested loops. The structure of the most computationally intensive loops is too complex to invert automatically or by hand. The only practical solution is to transpose the arrays so that the dimension accessed by the innermost loop is in cache order. The arrays can be transposed at construction or prior to entering a computationally intensive section of code. The former requires all array references to be modified, while the latter is cost effective only if the cost of the transpose is amortized over many accesses. I used the second approach to optimize code 10. Code 5 has four-dimensional arrays and loops are nested four deep. For all of the reasons cited above the compiler is not able to restructure three key loops. Assume C arrays and let the four dimensions of the arrays be i, j, k, and l. In the original code, the index structure of the three loops is L1: for i L2: for i L3: for i for l for l for j for k for j for k for j for k for l So only L3 accesses array elements in cache order. L1 is a very complex loop—much too complex to invert. I brought the loop into cache alignment by transposing the second and fourth dimensions of the arrays. Since the code uses a macro to compute all array indexes, I effected the transpose at construction and changed the macro appropriately. The dimensions of the new arrays are now: i, l, k, and j. L3 is a simple loop and easily inverted. L2 has a loop-carried scalar dependence in k. By promoting the scalar name that carries the dependence to an array, I was able to invert the third and fourth subloops aligning the loop with cache. Code 5 is by far the most difficult of the four codes to optimize for array accesses; but the knowledge required to fix the problems is no more than that required for the other codes. I would judge this code at the limits of, but not beyond, the capabilities of appropriately trained computational scientists. Array Strides When a cache miss occurs, a line (64 bytes) rather than just one word is loaded into the cache. If data is accessed stride 1, than the cost of the miss is amortized over 8 words. Any stride other than one reduces the cost savings. Two of the ten codes studied suffered from non-unit strides. The codes represent two important classes of "strided" codes. Code 1 employs a multi-grid algorithm to reduce time to convergence. The grids are every tenth, fifth, second, and unit element. Since time to convergence is inversely proportional to the distance between elements, coarse grids converge quickly providing good starting values for finer grids. The better starting values further reduce the time to convergence. The downside is that grids of every nth element, n > 1, introduce non-unit strides into the computation. In the original code, much of the savings of the multi-grid algorithm were lost due to this problem. I eliminated the problem by compressing (copying) coarse grids into continuous memory, and rewriting the computation as a function of the compressed grid. On convergence, I copied the final values of the compressed grid back to the original grid. The savings gained from unit stride access of the compressed grid more than paid for the cost of copying. Using compressed grids, the loop from code 1 included in the previous section becomes do j = 1, GZ do i = 1, GZ T1 = CA(i+0, j-1) + CA(i-1, j+0) T4 = CA1(i+1, j+0) + CA1(i+0, j+1) S1 = T1 + T4 - 4 * CA1(i+0, j+0) CA(i+0, j+0) = CA1(i+0, j+0) + DD * S1 enddo enddo where CA and CA1 are compressed arrays of size GZ. Code 7 traverses a list of objects selecting objects for later processing. The labels of the selected objects are stored in an array. The selection step has unit stride, but the processing steps have irregular stride. A fix is to save the parameters of the selected objects in temporary arrays as they are selected, and pass the temporary arrays to the processing functions. The fix is practical if the same parameters are used in selection as in processing, or if processing comprises a series of distinct steps which use overlapping subsets of the parameters. Both conditions are true for code 7, so I achieved significant improvement by copying parameters to temporary arrays during selection. Data reuse In the previous sections, we optimized for spatial locality. It is also important to optimize for temporal locality. Once read, a datum should be used as much as possible before it is forced from cache. Loop fusion and loop unrolling are two techniques that increase temporal locality. Unfortunately, both techniques increase register pressure—as loop bodies become larger, the number of registers required to hold temporary values grows. Once register spilling occurs, any gains evaporate quickly. For multiprocessors with small register sets or small caches, the sweet spot can be very small. In the ten codes presented here, I found no opportunities for loop fusion and only two opportunities for loop unrolling (codes 1 and 3). In code 1, unrolling the outer and inner loop one iteration increases the number of result values computed by the loop body from 1 to 4, do J = 1, GZ-2, 2 do I = 1, GZ-2, 2 T1 = CA1(i+0, j-1) + CA1(i-1, j+0) T2 = CA1(i+1, j-1) + CA1(i+0, j+0) T3 = CA1(i+0, j+0) + CA1(i-1, j+1) T4 = CA1(i+1, j+0) + CA1(i+0, j+1) T5 = CA1(i+2, j+0) + CA1(i+1, j+1) T6 = CA1(i+1, j+1) + CA1(i+0, j+2) T7 = CA1(i+2, j+1) + CA1(i+1, j+2) S1 = T1 + T4 - 4 * CA1(i+0, j+0) S2 = T2 + T5 - 4 * CA1(i+1, j+0) S3 = T3 + T6 - 4 * CA1(i+0, j+1) S4 = T4 + T7 - 4 * CA1(i+1, j+1) CA(i+0, j+0) = CA1(i+0, j+0) + DD * S1 CA(i+1, j+0) = CA1(i+1, j+0) + DD * S2 CA(i+0, j+1) = CA1(i+0, j+1) + DD * S3 CA(i+1, j+1) = CA1(i+1, j+1) + DD * S4 enddo enddo The loop body executes 12 reads, whereas as the rolled loop shown in the previous section executes 20 reads to compute the same four values. In code 3, two loops are unrolled 8 times and one loop is unrolled 4 times. Here is the before for (k = 0; k < NK[u]; k++) { sum = 0.0; for (y = 0; y < NY; y++) { sum += W[y][u][k] * delta[y]; } backprop[i++]=sum; } and after code for (k = 0; k < KK - 8; k+=8) { sum0 = 0.0; sum1 = 0.0; sum2 = 0.0; sum3 = 0.0; sum4 = 0.0; sum5 = 0.0; sum6 = 0.0; sum7 = 0.0; for (y = 0; y < NY; y++) { sum0 += W[y][0][k+0] * delta[y]; sum1 += W[y][0][k+1] * delta[y]; sum2 += W[y][0][k+2] * delta[y]; sum3 += W[y][0][k+3] * delta[y]; sum4 += W[y][0][k+4] * delta[y]; sum5 += W[y][0][k+5] * delta[y]; sum6 += W[y][0][k+6] * delta[y]; sum7 += W[y][0][k+7] * delta[y]; } backprop[k+0] = sum0; backprop[k+1] = sum1; backprop[k+2] = sum2; backprop[k+3] = sum3; backprop[k+4] = sum4; backprop[k+5] = sum5; backprop[k+6] = sum6; backprop[k+7] = sum7; } for one of the loops unrolled 8 times. Optimizing for temporal locality is the most difficult optimization considered in this paper. The concepts are not difficult, but the sweet spot is small. Identifying where the program can benefit from loop unrolling or loop fusion is not trivial. Moreover, it takes some effort to get it right. Still, educating scientific programmers about temporal locality and teaching them how to optimize for it will pay dividends. Reducing instruction count Execution time is a function of instruction count. Reduce the count and you usually reduce the time. The best solution is to use a more efficient algorithm; that is, an algorithm whose order of complexity is smaller, that converges quicker, or is more accurate. Optimizing source code without changing the algorithm yields smaller, but still significant, gains. This paper considers only the latter because the intent is to study how much better codes can run if written by programmers schooled in basic code optimization techniques. The ten codes studied benefited from three types of "instruction reducing" optimizations. The two most prevalent were hoisting invariant memory and data operations out of inner loops. The third was eliminating unnecessary data copying. The nature of these inefficiencies is language dependent. Memory operations The semantics of C make it difficult for the compiler to determine all the invariant memory operations in a loop. The problem is particularly acute for loops in functions since the compiler may not know the values of the function's parameters at every call site when compiling the function. Most compilers support pragmas to help resolve ambiguities; however, these pragmas are not comprehensive and there is no standard syntax. To guarantee that invariant memory operations are not executed repetitively, the user has little choice but to hoist the operations by hand. The problem is not as severe in Fortran programs because in the absence of equivalence statements, it is a violation of the language's semantics for two names to share memory. Codes 3 and 5 are C programs. In both cases, the compiler did not hoist all invariant memory operations from inner loops. Consider the following loop from code 3 for (y = 0; y < NY; y++) { i = 0; for (u = 0; u < NU; u++) { for (k = 0; k < NK[u]; k++) { dW[y][u][k] += delta[y] * I1[i++]; } } } Since dW[y][u] can point to the same memory space as delta for one or more values of y and u, assignment to dW[y][u][k] may change the value of delta[y]. In reality, dW and delta do not overlap in memory, so I rewrote the loop as for (y = 0; y < NY; y++) { i = 0; Dy = delta[y]; for (u = 0; u < NU; u++) { for (k = 0; k < NK[u]; k++) { dW[y][u][k] += Dy * I1[i++]; } } } Failure to hoist invariant memory operations may be due to complex address calculations. If the compiler can not determine that the address calculation is invariant, then it can hoist neither the calculation nor the associated memory operations. As noted above, code 5 uses a macro to address four-dimensional arrays #define MAT4D(a,q,i,j,k) (double *)((a)->data + (q)*(a)->strides[0] + (i)*(a)->strides[3] + (j)*(a)->strides[2] + (k)*(a)->strides[1]) The macro is too complex for the compiler to understand and so, it does not identify any subexpressions as loop invariant. The simplest way to eliminate the address calculation from the innermost loop (over i) is to define a0 = MAT4D(a,q,0,j,k) before the loop and then replace all instances of *MAT4D(a,q,i,j,k) in the loop with a0[i] A similar problem appears in code 6, a Fortran program. The key loop in this program is do n1 = 1, nh nx1 = (n1 - 1) / nz + 1 nz1 = n1 - nz * (nx1 - 1) do n2 = 1, nh nx2 = (n2 - 1) / nz + 1 nz2 = n2 - nz * (nx2 - 1) ndx = nx2 - nx1 ndy = nz2 - nz1 gxx = grn(1,ndx,ndy) gyy = grn(2,ndx,ndy) gxy = grn(3,ndx,ndy) balance(n1,1) = balance(n1,1) + (force(n2,1) * gxx + force(n2,2) * gxy) * h1 balance(n1,2) = balance(n1,2) + (force(n2,1) * gxy + force(n2,2) * gyy)*h1 end do end do The programmer has written this loop well—there are no loop invariant operations with respect to n1 and n2. However, the loop resides within an iterative loop over time and the index calculations are independent with respect to time. Trading space for time, I precomputed the index values prior to the entering the time loop and stored the values in two arrays. I then replaced the index calculations with reads of the arrays. Data operations Ways to reduce data operations can appear in many forms. Implementing a more efficient algorithm produces the biggest gains. The closest I came to an algorithm change was in code 4. This code computes the inner product of K-vectors A(i) and B(j), 0 = i < N, 0 = j < M, for most values of i and j. Since the program computes most of the NM possible inner products, it is more efficient to compute all the inner products in one triply-nested loop rather than one at a time when needed. The savings accrue from reading A(i) once for all B(j) vectors and from loop unrolling. for (i = 0; i < N; i+=8) { for (j = 0; j < M; j++) { sum0 = 0.0; sum1 = 0.0; sum2 = 0.0; sum3 = 0.0; sum4 = 0.0; sum5 = 0.0; sum6 = 0.0; sum7 = 0.0; for (k = 0; k < K; k++) { sum0 += A[i+0][k] * B[j][k]; sum1 += A[i+1][k] * B[j][k]; sum2 += A[i+2][k] * B[j][k]; sum3 += A[i+3][k] * B[j][k]; sum4 += A[i+4][k] * B[j][k]; sum5 += A[i+5][k] * B[j][k]; sum6 += A[i+6][k] * B[j][k]; sum7 += A[i+7][k] * B[j][k]; } C[i+0][j] = sum0; C[i+1][j] = sum1; C[i+2][j] = sum2; C[i+3][j] = sum3; C[i+4][j] = sum4; C[i+5][j] = sum5; C[i+6][j] = sum6; C[i+7][j] = sum7; }} This change requires knowledge of a typical run; i.e., that most inner products are computed. The reasons for the change, however, derive from basic optimization concepts. It is the type of change easily made at development time by a knowledgeable programmer. In code 5, we have the data version of the index optimization in code 6. Here a very expensive computation is a function of the loop indices and so cannot be hoisted out of the loop; however, the computation is invariant with respect to an outer iterative loop over time. We can compute its value for each iteration of the computation loop prior to entering the time loop and save the values in an array. The increase in memory required to store the values is small in comparison to the large savings in time. The main loop in Code 8 is doubly nested. The inner loop includes a series of guarded computations; some are a function of the inner loop index but not the outer loop index while others are a function of the outer loop index but not the inner loop index for (j = 0; j < N; j++) { for (i = 0; i < M; i++) { r = i * hrmax; R = A[j]; temp = (PRM[3] == 0.0) ? 1.0 : pow(r, PRM[3]); high = temp * kcoeff * B[j] * PRM[2] * PRM[4]; low = high * PRM[6] * PRM[6] / (1.0 + pow(PRM[4] * PRM[6], 2.0)); kap = (R > PRM[6]) ? high * R * R / (1.0 + pow(PRM[4]*r, 2.0) : low * pow(R/PRM[6], PRM[5]); < rest of loop omitted > }} Note that the value of temp is invariant to j. Thus, we can hoist the computation for temp out of the loop and save its values in an array. for (i = 0; i < M; i++) { r = i * hrmax; TEMP[i] = pow(r, PRM[3]); } [N.B. – the case for PRM[3] = 0 is omitted and will be reintroduced later.] We now hoist out of the inner loop the computations invariant to i. Since the conditional guarding the value of kap is invariant to i, it behooves us to hoist the computation out of the inner loop, thereby executing the guard once rather than M times. The final version of the code is for (j = 0; j < N; j++) { R = rig[j] / 1000.; tmp1 = kcoeff * par[2] * beta[j] * par[4]; tmp2 = 1.0 + (par[4] * par[4] * par[6] * par[6]); tmp3 = 1.0 + (par[4] * par[4] * R * R); tmp4 = par[6] * par[6] / tmp2; tmp5 = R * R / tmp3; tmp6 = pow(R / par[6], par[5]); if ((par[3] == 0.0) && (R > par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * tmp5; } else if ((par[3] == 0.0) && (R <= par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * tmp4 * tmp6; } else if ((par[3] != 0.0) && (R > par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * TEMP[i] * tmp5; } else if ((par[3] != 0.0) && (R <= par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * TEMP[i] * tmp4 * tmp6; } for (i = 0; i < M; i++) { kap = KAP[i]; r = i * hrmax; < rest of loop omitted > } } Maybe not the prettiest piece of code, but certainly much more efficient than the original loop, Copy operations Several programs unnecessarily copy data from one data structure to another. This problem occurs in both Fortran and C programs, although it manifests itself differently in the two languages. Code 1 declares two arrays—one for old values and one for new values. At the end of each iteration, the array of new values is copied to the array of old values to reset the data structures for the next iteration. This problem occurs in Fortran programs not included in this study and in both Fortran 77 and Fortran 90 code. Introducing pointers to the arrays and swapping pointer values is an obvious way to eliminate the copying; but pointers is not a feature that many Fortran programmers know well or are comfortable using. An easy solution not involving pointers is to extend the dimension of the value array by 1 and use the last dimension to differentiate between arrays at different times. For example, if the data space is N x N, declare the array (N, N, 2). Then store the problem’s initial values in (_, _, 2) and define the scalar names new = 2 and old = 1. At the start of each iteration, swap old and new to reset the arrays. The old–new copy problem did not appear in any C program. In programs that had new and old values, the code swapped pointers to reset data structures. Where unnecessary coping did occur is in structure assignment and parameter passing. Structures in C are handled much like scalars. Assignment causes the data space of the right-hand name to be copied to the data space of the left-hand name. Similarly, when a structure is passed to a function, the data space of the actual parameter is copied to the data space of the formal parameter. If the structure is large and the assignment or function call is in an inner loop, then copying costs can grow quite large. While none of the ten programs considered here manifested this problem, it did occur in programs not included in the study. A simple fix is always to refer to structures via pointers. Optimizing loop structures Since scientific programs spend almost all their time in loops, efficient loops are the key to good performance. Conditionals, function calls, little instruction level parallelism, and large numbers of temporary values make it difficult for the compiler to generate tightly packed, highly efficient code. Conditionals and function calls introduce jumps that disrupt code flow. Users should eliminate or isolate conditionls to their own loops as much as possible. Often logical expressions can be substituted for if-then-else statements. For example, code 2 includes the following snippet MaxDelta = 0.0 do J = 1, N do I = 1, M < code omitted > Delta = abs(OldValue ? NewValue) if (Delta > MaxDelta) MaxDelta = Delta enddo enddo if (MaxDelta .gt. 0.001) goto 200 Since the only use of MaxDelta is to control the jump to 200 and all that matters is whether or not it is greater than 0.001, I made MaxDelta a boolean and rewrote the snippet as MaxDelta = .false. do J = 1, N do I = 1, M < code omitted > Delta = abs(OldValue ? NewValue) MaxDelta = MaxDelta .or. (Delta .gt. 0.001) enddo enddo if (MaxDelta) goto 200 thereby, eliminating the conditional expression from the inner loop. A microprocessor can execute many instructions per instruction cycle. Typically, it can execute one or more memory, floating point, integer, and jump operations. To be executed simultaneously, the operations must be independent. Thick loops tend to have more instruction level parallelism than thin loops. Moreover, they reduce memory traffice by maximizing data reuse. Loop unrolling and loop fusion are two techniques to increase the size of loop bodies. Several of the codes studied benefitted from loop unrolling, but none benefitted from loop fusion. This observation is not too surpising since it is the general tendency of programmers to write thick loops. As loops become thicker, the number of temporary values grows, increasing register pressure. If registers spill, then memory traffic increases and code flow is disrupted. A thick loop with many temporary values may execute slower than an equivalent series of thin loops. The biggest gain will be achieved if the thick loop can be split into a series of independent loops eliminating the need to write and read temporary arrays. I found such an occasion in code 10 where I split the loop do i = 1, n do j = 1, m A24(j,i)= S24(j,i) * T24(j,i) + S25(j,i) * U25(j,i) B24(j,i)= S24(j,i) * T25(j,i) + S25(j,i) * U24(j,i) A25(j,i)= S24(j,i) * C24(j,i) + S25(j,i) * V24(j,i) B25(j,i)= S24(j,i) * U25(j,i) + S25(j,i) * V25(j,i) C24(j,i)= S26(j,i) * T26(j,i) + S27(j,i) * U26(j,i) D24(j,i)= S26(j,i) * T27(j,i) + S27(j,i) * V26(j,i) C25(j,i)= S27(j,i) * S28(j,i) + S26(j,i) * U28(j,i) D25(j,i)= S27(j,i) * T28(j,i) + S26(j,i) * V28(j,i) end do end do into two disjoint loops do i = 1, n do j = 1, m A24(j,i)= S24(j,i) * T24(j,i) + S25(j,i) * U25(j,i) B24(j,i)= S24(j,i) * T25(j,i) + S25(j,i) * U24(j,i) A25(j,i)= S24(j,i) * C24(j,i) + S25(j,i) * V24(j,i) B25(j,i)= S24(j,i) * U25(j,i) + S25(j,i) * V25(j,i) end do end do do i = 1, n do j = 1, m C24(j,i)= S26(j,i) * T26(j,i) + S27(j,i) * U26(j,i) D24(j,i)= S26(j,i) * T27(j,i) + S27(j,i) * V26(j,i) C25(j,i)= S27(j,i) * S28(j,i) + S26(j,i) * U28(j,i) D25(j,i)= S27(j,i) * T28(j,i) + S26(j,i) * V28(j,i) end do end do Conclusions Over the course of the last year, I have had the opportunity to work with over two dozen academic scientific programmers at leading research universities. Their research interests span a broad range of scientific fields. Except for two programs that relied almost exclusively on library routines (matrix multiply and fast Fourier transform), I was able to improve significantly the single processor performance of all codes. Improvements range from 2x to 15.5x with a simple average of 4.75x. Changes to the source code were at a very high level. I did not use sophisticated techniques or programming tools to discover inefficiencies or effect the changes. Only one code was parallel despite the availability of parallel systems to all developers. Clearly, we have a problem—personal scientific research codes are highly inefficient and not running parallel. The developers are unaware of simple optimization techniques to make programs run faster. They lack education in the art of code optimization and parallel programming. I do not believe we can fix the problem by publishing additional books or training manuals. To date, the developers in questions have not studied the books or manual available, and are unlikely to do so in the future. Short courses are a possible solution, but I believe they are too concentrated to be much use. The general concepts can be taught in a three or four day course, but that is not enough time for students to practice what they learn and acquire the experience to apply and extend the concepts to their codes. Practice is the key to becoming proficient at optimization. I recommend that graduate students be required to take a semester length course in optimization and parallel programming. We would never give someone access to state-of-the-art scientific equipment costing hundreds of thousands of dollars without first requiring them to demonstrate that they know how to use the equipment. Yet the criterion for time on state-of-the-art supercomputers is at most an interesting project. Requestors are never asked to demonstrate that they know how to use the system, or can use the system effectively. A semester course would teach them the required skills. Government agencies that fund academic scientific research pay for most of the computer systems supporting scientific research as well as the development of most personal scientific codes. These agencies should require graduate schools to offer a course in optimization and parallel programming as a requirement for funding. About the Author John Feo received his Ph.D. in Computer Science from The University of Texas at Austin in 1986. After graduate school, Dr. Feo worked at Lawrence Livermore National Laboratory where he was the Group Leader of the Computer Research Group and principal investigator of the Sisal Language Project. In 1997, Dr. Feo joined Tera Computer Company where he was project manager for the MTA, and oversaw the programming and evaluation of the MTA at the San Diego Supercomputer Center. In 2000, Dr. Feo joined Sun Microsystems as an HPC application specialist. He works with university research groups to optimize and parallelize scientific codes. Dr. Feo has published over two dozen research articles in the areas of parallel parallel programming, parallel programming languages, and application performance.

  Read the article

• Building a SOA/BPM/BAM Cluster Part I &ndash; Preparing the Environment

  - by antony.reynolds

  An increasing number of customers are using SOA Suite in a cluster configuration, I might hazard to say that the majority of production deployments are now using SOA clusters.  So I thought it may be useful to detail the steps in building an 11g cluster and explain a little about why things are done the way they are. In this series of posts I will explain how to build a SOA/BPM cluster using the Enterprise Deployment Guide. This post will explain the setting required to prepare the cluster for installation and configuration. Software Required The following software is required for an 11.1.1.3 SOA/BPM install. Software Version Notes Oracle Database Certified databases are listed here SOA & BPM Suites require a working database installation. Repository Creation Utility (RCU) 11.1.1.3 If upgrading an 11.1.1.2 repository then a separate script is available. Web Tier Utilities 11.1.1.3 Provides Web Server, 11.1.1.3 is an upgrade to 11.1.1.2, so 11.1.1.2 must be installed first. Web Tier Utilities 11.1.1.3 Web Server, 11.1.1.3 Patch.  You can use the 11.1.1.2 version without problems. Oracle WebLogic Server 11gR1 10.3.3 This is the host platform for 11.1.1.3 SOA/BPM Suites. SOA Suite 11.1.1.2 SOA Suite 11.1.1.3 is an upgrade to 11.1.1.2, so 11.1.1.2 must be installed first. SOA Suite 11.1.1.3 SOA Suite 11.1.1.3 patch, requires 11.1.12 to have been installed. My installation was performed on Oracle Enterprise Linux 5.4 64-bit. Database I will not cover setting up the database in this series other than to identify the database requirements.  If setting up a SOA cluster then ideally we would also be using a RAC database.  I assume that this is running on separate machines to the SOA cluster.  Section 2.1, “Database”, of the EDG covers the database configuration in detail. Settings The database should have processes set to at least 400 if running SOA/BPM and BAM. alter system set processes=400 scope=spfile Run RCU The Repository Creation Utility creates the necessary database tables for the SOA Suite.  The RCU can be run from any machine that can access the target database.  In 11g the RCU creates a number of pre-defined users and schema with a user defiend prefix.  This allows you to have multiple 11g installations in the same database. After running the RCU you need to grant some additional privileges to the soainfra user.  The soainfra user should have privileges on the transaction tables. grant select on sys.dba_pending_transactions to prefix_soainfra Grant force any transaction to prefix_soainfra Machines The cluster will be built on the following machines. EDG Name is the name used for this machine in the EDG. Notes are a description of the purpose of the machine. EDG Name Notes LB External load balancer to distribute load across and failover between web servers. WEBHOST1 Hosts a web server. WEBHOST2 Hosts a web server. SOAHOST1 Hosts SOA components. SOAHOST2 Hosts SOA components. BAMHOST1 Hosts BAM components. BAMHOST2 Hosts BAM components. Note that it is possible to collapse the BAM servers so that they run on the same machines as the SOA servers. In this case BAMHOST1 and SOAHOST1 would be the same, as would BAMHOST2 and SOAHOST2. The cluster may include more than 2 servers and in this case we add SOAHOST3, SOAHOST4 etc as needed. My cluster has WEBHOST1, SOAHOST1 and BAMHOST1 all running on a single machine. Software Components The cluster will use the following software components. EDG Name is the name used for this machine in the EDG. Type is the type of component, generally a WebLogic component. Notes are a description of the purpose of the component. EDG Name Type Notes AdminServer Admin Server Domain Admin Server WLS_WSM1 Managed Server Web Services Manager Policy Manager Server WLS_WSM2 Managed Server Web Services Manager Policy Manager Server WLS_SOA1 Managed Server SOA/BPM Managed Server WLS_SOA2 Managed Server SOA/BPM Managed Server WLS_BAM1 Managed Server BAM Managed Server running Active Data Cache WLS_BAM2 Managed Server BAM Manager Server without Active Data Cache   Node Manager Will run on all hosts with WLS servers OHS1 Web Server Oracle HTTP Server OHS2 Web Server Oracle HTTP Server LB Load Balancer Load Balancer, not part of SOA Suite The above assumes a 2 node cluster. Network Configuration The SOA cluster requires an extensive amount of network configuration.  I would recommend assigning a private sub-net (internal IP addresses such as 10.x.x.x, 192.168.x.x or 172.168.x.x) to the cluster for use by addresses that only need to be accessible to the Load Balancer or other cluster members.  Section 2.2, "Network", of the EDG covers the network configuration in detail. EDG Name is the hostname used in the EDG. IP Name is the IP address name used in the EDG. Type is the type of IP address: Fixed is fixed to a single machine. Floating is assigned to one of several machines to allow for server migration. Virtual is assigned to a load balancer and used to distribute load across several machines. Host is the host where this IP address is active.  Note for floating IP addresses a range of hosts is given. Bound By identifies which software component will use this IP address. Scope shows where this IP address needs to be resolved. Cluster scope addresses only have to be resolvable by machines in the cluster, i.e. the machines listed in the previous section.  These addresses are only used for inter-cluster communication or for access by the load balancer. Internal scope addresses Notes are comments on why that type of IP is used. EDG Name IP Name Type Host Bound By Scope Notes ADMINVHN VIP1 Floating SOAHOST1-SOAHOSTn AdminServer Cluster Admin server, must be able to migrate between SOA server machines. SOAHOST1 IP1 Fixed SOAHOST1 NodeManager, WLS_WSM1 Cluster WSM Server 1 does not require server migration. SOAHOST2 IP2 Fixed SOAHOST1 NodeManager, WLS_WSM2 Cluster WSM Server 2 does not require server migration SOAHOST1VHN VIP2 Floating SOAHOST1-SOAHOSTn WLS_SOA1 Cluster SOA server 1, must be able to migrate between SOA server machines SOAHOST2VHN VIP3 Floating SOAHOST1-SOAHOSTn WLS_SOA2 Cluster SOA server 2, must be able to migrate between SOA server machines BAMHOST1 IP4 Fixed BAMHOST1 NodeManager Cluster   BAMHOST1VHN VIP4 Floating BAMHOST1-BAMHOSTn WLS_BAM1 Cluster BAM server 1, must be able to migrate between BAM server machines BAMHOST2 IP3 Fixed BAMHOST2 NodeManager, WLS_BAM2 Cluster BAM server 2 does not require server migration WEBHOST1 IP5 Fixed WEBHOST1 OHS1 Cluster   WEBHOST2 IP6 Fixed WEBHOST2 OHS2 Cluster   soa.mycompany.com VIP5 Virtual LB LB Public External access point to SOA cluster. admin.mycompany.com VIP6 Virtual LB LB Internal Internal access to WLS console and EM soainternal.mycompany.com VIP7 Virtual LB LB Internal Internal access point to SOA cluster Floating IP addresses are IP addresses that may be re-assigned between machines in the cluster.  For example in the event of failure of SOAHOST1 then WLS_SOA1 will need to be migrated to another server.  In this case VIP2 (SOAHOST1VHN) will need to be activated on the new target machine.  Once set up the node manager will manage registration and removal of the floating IP addresses with the exception of the AdminServer floating IP address. Note that if the BAMHOSTs and SOAHOSTs are the same machine then you can obviously share the hostname and fixed IP addresses, but you still need separate floating IP addresses for the different managed servers.  The hostnames don’t have to be the ones given in the EDG, but they must be distinct in the same way as the ETC names are distinct.  If the type is a fixed IP then if the addresses are the same you can use the same hostname, for example if you collapse the soahost1, bamhost1 and webhost1 onto a single machine then you could refer to them all as HOST1 and give them the same IP address, however SOAHOST1VHN can never be the same as BAMHOST1VHN because these are floating IP addresses. Notes on DNS IP addresses that are of scope “Cluster” just need to be in the hosts file (/etc/hosts on Linux, C:\Windows\System32\drivers\etc\hosts on Windows) of all the machines in the cluster and the load balancer.  IP addresses that are of scope “Internal” need to be available on the internal DNS servers, whilst IP addresses of scope “Public” need to be available on external and internal DNS servers. Shared File System At a minimum the cluster needs shared storage for the domain configuration, XA transaction logs and JMS file stores.  It is also possible to place the software itself on a shared server.  I strongly recommend that all machines have the same file structure for their SOA installation otherwise you will experience pain!  Section 2.3, "Shared Storage and Recommended Directory Structure", of the EDG covers the shared storage recommendations in detail. The following shorthand is used for locations: ORACLE_BASE is the root of the file system used for software and configuration files. MW_HOME is the location used by the installed SOA/BPM Suite installation.  This is also used by the web server installation.  In my installation it is set to <ORACLE_BASE>/SOA11gPS2. ORACLE_HOME is the location of the Oracle SOA components or the Oracle Web components.  This directory is installed under the the MW_HOME but the name is decided by the user at installation, default values are Oracle_SOA1 and Oracle_Web1.  In my installation they are set to <MW_HOME>/Oracle_SOA and <MW_HOME>/Oracle _WEB. ORACLE_COMMON_HOME is the location of the common components and is located under the MW_HOME directory.  This is always <MW_HOME>/oracle_common. ORACLE_INSTANCE is used by the Oracle HTTP Server and/or Oracle Web Cache.  It is recommended to create it under <ORACLE_BASE>/admin.  In my installation they are set to <ORACLE_BASE>/admin/Web1, <ORACLE_BASE>/admin/Web2 and <ORACLE_BASE>/admin/WC1. WL_HOME is the WebLogic server home and is always found at <MW_HOME>/wlserver_10.3. Key file locations are shown below. Directory Notes <ORACLE_BASE>/admin/domain_name/aserver/domain_name Shared location for domain.  Used to allow admin server to manually fail over between machines.  When creating domain_name provide the aserver directory as the location for the domain. In my install this is <ORACLE_BASE>/admin/aserver/soa_domain as I only have one domain on the box. <ORACLE_BASE>/admin/domain_name/aserver/applications Shared location for deployed applications.  Needs to be provided when creating the domain. In my install this is <ORACLE_BASE>/admin/aserver/applications as I only have one domain on the box. <ORACLE_BASE>/admin/domain_name/mserver/domain_name Either unique location for each machine or can be shared between machines to simplify task of packing and unpacking domain.  This acts as the managed server configuration location.  Keeping it separate from Admin server helps to avoid problems with the managed servers messing up the Admin Server. In my install this is <ORACLE_BASE>/admin/mserver/soa_domain as I only have one domain on the box. <ORACLE_BASE>/admin/domain_name/mserver/applications Either unique location for each machine or can be shared between machines.  Holds deployed applications. In my install this is <ORACLE_BASE>/admin/mserver/applications as I only have one domain on the box. <ORACLE_BASE>/admin/domain_name/soa_cluster_name Shared directory to hold the following   dd – deployment descriptors   jms – shared JMS file stores   fadapter – shared file adapter co-ordination files   tlogs – shared transaction log files In my install this is <ORACLE_BASE>/admin/soa_cluster. <ORACLE_BASE>/admin/instance_name Local folder for web server (OHS) instance. In my install this is <ORACLE_BASE>/admin/web1 and <ORACLE_BASE>/admin/web2. I also have <ORACLE_BASE>/admin/wc1 for the Web Cache I use as a load balancer. <ORACLE_BASE>/product/fmw This can be a shared or local folder for the SOA/BPM Suite software.  I used a shared location so I only ran the installer once. In my install this is <ORACLE_BASE>/SOA11gPS2 All the shared files need to be put onto a shared storage media.  I am using NFS, but recommendation for production would be a SAN, with mirrored disks for resilience. Collapsing Environments To reduce the hardware requirements it is possible to collapse the BAMHOST, SOAHOST and WEBHOST machines onto a single physical machine.  This will require more memory but memory is a lot cheaper than additional machines.  For environments that require higher security then stay with a separate WEBHOST tier as per the EDG.  Similarly for high volume environments then keep a separate set of machines for BAM and/or Web tier as per the EDG. Notes on Dev Environments In a dev environment it is acceptable to use a a single node (non-RAC) database, but be aware that the config of the data sources is different (no need to use multi-data source in WLS).  Typically in a dev environment we will collapse the BAMHOST, SOAHOST and WEBHOST onto a single machine and use a software load balancer.  To test a cluster properly we will need at least 2 machines. For my test environment I used Oracle Web Cache as a load balancer.  I ran it on one of the SOA Suite machines and it load balanced across the Web Servers on both machines.  This was easy for me to set up and I could administer it from a web based console.

  Read the article

• Prepping a conference

  - by Laurent Bugnion

  I have had the chance to talk at many conferences these past few years, and came up with a way to prepare them which works really well for me. Most importantly, it would make it quite easy to overcome an emergency (for example if my laptop would suddenly lose data). The whole code as well as the slides and other documents are in the cloud. I also use source control for my demos, so that I always have the latest and the greatest, but also a history of changes I made to my demos. Finally I have a system of code snippets which works great, and I often had very positive remarks from the audience regarding that. Putting everything in the cloud The one thing I used to be the most scared of was a sudden crash of my laptop, and being unable to restore in time for a conference. Most conferences ask speakers to send slides a few days (or weeks…) in advance, but let's face it, we all have last minute changes to our talks and I always come in the conference with updated slides that I pass to the management team. The answer to that dilemma used to be working off memory sticks, and that worked not bad. However last year I started putting all the documents relating to a conference in a DropBox folder, and that works great too. Obviously DropBox works only if you have connectivity, so if I for instance update slides while on an international flight, I cannot save to the cloud. The obvious answer to that is to backup everything on a memory stick… but I have to admit, I have been trusting my luck and working off my laptop HD and then synching everything to the cloud after landing. Of course on some US national flights you get WiFi on board, so in that case it is even simpler :) Usually after the conference is done, I remove the files from DropBox and copy them to their "final destination". They are backed up from there to BackBlaze, the great online backup service I am using routinely (I currently have about 90GB of data in BackBlaze). Outlining the presentations I like to have a written outline of my presentations written somewhere. I keep it simple, just write the various sections of the presentation with timing. I guess it is a remnant of the time when I was a private pilot, and using checklists for flight preparation. For example: Demo about designability 15' (0:37) Switch to Blend Open MainPage.xaml Create a DataTemplate ... Here I can immediately see during the presentation if I am taking too much time for my demo (0:37 is where I need to be when I am done with this section of the presentation, and 15' is the time that this particular section takes). I keep these sections reasonable, I don't detail every step of the preparation. Typically I have one such section for every 10-15 minutes of my talks. Yes, I am timing my presentations. I keep adjusting these numbers when I rehearse, and this really helps to feel more confident during the presentations. This is especially important for presentations that are long, like my MIX11 demo which clocked at 57 minutes (I had a lot of stuff to show…). Such presentations are risky, because if anything goes wrong, you will have to cut stuff, so the answer to that is: Rehearse, rehearse and when you're done rehearsing, rehearse a little more. I also have a "Preparation" section where I outline what I need to do before a presentation. For instance: Preparation Reboot in VHD Make sure MSN and Twitter are not running. Open VS10 and load demo Open Blend and load demo Run the WP7 emulator ... I typically start preparing my laptop an hour before the talk, starting everything I need to start and then putting my laptop to sleep. Saving and printing the outline, Timing Printing is a real problem because it is really hard to find a printer at most conference venues, and also quite hard in hotels. To solve that, I simply write everything in OneNote (synched to the cloud, now you start to know what I like ;) and then I print it to a PDF (I use CutePDFWriter) that I save to my Kindle. During the presentation, I read the outline off the Kindle (I mostly just need a quick check to see how I am timing). For timing during the presentation, I use the free tool ChronoGPS on my Windows Phone 7, but of course any phone these days has a clock/chrono application. In some conferences, they even have timers that the presenters can see, but they tend to count down and I prefer to count up… so I just use my own :) Source control for demos For demos, I create a separate folder and use Mercurial as source control. Mercurial has the huge advantage (over SVN or TFS) to work offline too, so I can commit while on a plane, and all the history is saved. Then when I have connectivity I push everything to the cloud (I am using the fantastic Trunksapp.com for my private repositories). Here too the obvious downside is the risk of losing my last changes if my laptop crashes before I can push to the cloud, and here too the obvious answer would be to work from a memory stick… though I have to admit I didn't do that lately (except when I was writing Silverlight 4 Unleashed, where I was really paranoid…) And code snippets? I am one of these presenters who hates to type in front of an audience. I can type really fast (writing two books has this advantage, it really teaches you to touch type and be fast at it) but in the context of an audience, on a stage where it is often damn cold (an issue I had a lot in past conferences, air conditioning can freeze your fingers and make it really hard to type), it doesn't work as well. I don't know for you, but I really dislike seeing a presentation where the speaker uses the backspace key more often than others ;) To solve that, I like to have my code ready in snippets, and drag them to the screen. Then I can spend time explaining each code snippet, while highlighting portions of the code (always highlight what you talk about, the audience often doesn't even see the cursor and doesn't know where you are on the screen!) Over the years I have used various solutions for code snippets, and now I have one which works really well… if you take a few precautions! I use the Visual Studio Toolbox. Preparing the code snippets You can store code snippets in the Toolbox for anything, XAML, C# etc. I arrange the snippets in the order in which I need them, which is a great way to remember what comes next in the presentation. I also separate them by topic, to make it easier to find them, for example when I switch to the slides and then back to the code. Remember that no matter how experienced you are, you will feel more nervous on stage than while you are preparing, so any way to make it easier for you is going to be beneficial to the audience. To store a code snippet, I do the following: Open the final demo that you want to show to the audience in Visual Studio. In your code, select a snippet of code that you want to explain in particular. Make sure that the Visual Studio Toolbox is open (menu View, Toolbox or Ctrl-Alt-X). Drag the selected snippet from the code window to the toolbox. (if needed) drag the snippet to the correct location (for example between two other code snippets so that you can access it as you speak through the demo). Right click on the snippet and select Rename Item from the context menu. Select a meaningful name. For me I use the following conventions: If it is a method, I use the method's name. If it is not a whole method, I use a descriptive name. If it is the content of a method (i.e. the body only, without the method's signature), I use "-> MethodName". This reminds me during the presentation that this is only the body, and that I need to insert that into an existing signature. This is the case, for instance, when I use Visual Studio to automatically generate the members of an interface’s implementation; then I only need to insert my snippet inside the generated method body. Saving the snippets This is the most important!! It happened to me a few times that VS10 lost its settings. When that happens, the snippets are lost too! Yeah that really sucks, especially (as it happened once) when this is the case about an hour before a talk… Stress and sweat follows, not good conditions to start a talk in front of an audience believe me. Thankfully, saving snippets is really easy with the following steps: Select the menu Tools, Import and Export Settings. Select Export selected environment settings and press Next. Uncheck All Settings. Then expand General Settings and select Toolbox (only!). Press Next. Select your source control folder and save under a meaningful name (for instance Snippets.vssettings). Commit to source control and push to the cloud. By the way, this also has the advantage of applying source control to the snippets file (which is an XML file), so you get history for free on that file! Reimporting the snippets If VS loses its settings and you need to reimport the snippets, this can be done super easily and very fast. Make sure that the Toolbox is empty. When you import snippets, they are merged with existing ones, they do not replace the content of the Toolbox. Unless merging is really what you want, make sure that your Toolbox is clean before you import, it is really easier. Select the menu Tools, Import and Export Settings. Select Import selected environment settings and press Next. Select No, just import new settings and press Next. Press Browse and select the Snippets.vssettings file. Press Finish. Et voila, all your snippets appear again in the Toolbox. Whew, the worst was averted and you can start your demo without sweating! (I had to do that once literally 5 minutes before the start of a demo, while my laptop was already hooked to the projector, and it went just fine). What about special tools? When using special tools (for example beta versions of tools you have an early access to), or a special configuration of your laptop, things can get tricky because you cannot really be sure that you will get a laptop with the same tools and the same configuration at the conference. To solve that, I use the following precautions: I make my demos from a Virtual Hard Disk. The great John Papa made a very easy-to-follow web page where he explains how to create a VHD and install Win7 to it. This gives you the full power of your laptop (as fast as booting from the metal). For me, I have a basic configuration that I saved on a USB harddrive (Win7 plus drivers, basic settings for desktop, folder options, taskbar etc) and Visual Studio 2010 SP1 on it. When preparing, I start by copying this "basis VHD" to my laptop. I install additional tools and configurations. I save the VHD back to the USB harddrive in a different folder. This would allow me to reinstall my demo environment quite fast, for example in case of harddrive failure. Replace the harddrive, copy the VHD to it, configure the BCD and you can start. Unfortunately this only works if the laptop itself still works. In the worst case of total failure, my security is to back all the installers up: The installers I use are synched on all my laptops and backed up to BackBlaze. If the worst happens and my laptop is absolutely broken, I can download the installer from BackBlaze and install on another laptop. This of course takes some time, and if that happens 5 minutes before a presentation, well… I don't have an answer to that, except of course crossing my fingers. Still, all that gives me additional security. Conclusion Remember folks, talking to an audience, large or small, will make you nervous. Just ask Scott Hanselman :) The goal here is to create the best possible conditions for you, and to create an environment where everything is saved and easy to restore, where everything is well known and provides you with additional confidence. The cooler you feel before the presentation (and during ;)), the better your presentation will be. Here too, the goal is to provide the best user experience you can have, which in turn will make it more enjoyable for your audience! Happy presenting :) Laurent   Laurent Bugnion (GalaSoft) Subscribe | Twitter | Facebook | Flickr | LinkedIn

  Read the article

• Entity Framework 4.3.1 Code based Migrations and Connector/Net 6.6

  - by GABMARTINEZ

   Code-based migrations is a new feature as part of the Connector/Net support for Entity Framework 4.3.1. In this tutorial we'll see how we can use it so we can keep track of the changes done to our database creating a new application using the code first approach. If you don't have a clear idea about how code first works we highly recommend you to check this subject before going further with this tutorial. Creating our Model and Database with Code First  From VS 2010  1. Create a new console application 2.  Add the latest Entity Framework official package using Package Manager Console (Tools Menu, then Library Package Manager -> Package Manager Console). In the Package Manager Console we have to type  Install-Package EntityFramework This will add the latest version of this library.  We will also need to make some changes to your config file. A <configSections> was added which contains the version you have from EntityFramework.  An <entityFramework> section was also added where you can set up some initialization. This section is optional and by default is generated to use SQL Express. Since we don't need it for now (we'll see more about it below) let's leave this section empty as shown below. 3. Create a new Model with a simple entity. 4. Enable Migrations to generate the our Configuration class. In the Package Manager Console we have to type  Enable-Migrations; This will make some changes in our application. It will create a new folder called Migrations where all the migrations representing the changes we do to our model.  It will also create a Configuration class that we'll be using to initialize our SQL Generator and some other values like if we want to enable Automatic Migrations.  You can see that it already has the name of our DbContext. You can also create you Configuration class manually. 5. Specify our Model Provider. We need to specify in our Class Configuration that we'll be using MySQLClient since this is not part of the generated code. Also please make sure you have added the MySql.Data and the MySql.Data.Entity references to your project. using MySql.Data.Entity;   // Add the MySQL.Data.Entity namespace public Configuration() { this.AutomaticMigrationsEnabled = false; SetSqlGenerator("MySql.Data.MySqlClient", new MySql.Data.Entity.MySqlMigrationSqlGenerator());    // This will add our MySQLClient as SQL Generator } 6. Add our Data Provider and set up our connection string <connectionStrings> <add name="PersonalContext" connectionString="server=localhost;User Id=root;database=Personal;" providerName="MySql.Data.MySqlClient" /> </connectionStrings> <system.data> <DbProviderFactories> <remove invariant="MySql.Data.MySqlClient" /> <add name="MySQL Data Provider" invariant="MySql.Data.MySqlClient" description=".Net Framework Data Provider for MySQL" type="MySql.Data.MySqlClient.MySqlClientFactory, MySql.Data, Version=6.6.2.0, Culture=neutral, PublicKeyToken=c5687fc88969c44d" /> </DbProviderFactories> </system.data> * The version recommended to use of Connector/Net is 6.6.2 or earlier. At this point we can create our database and then start working with Migrations. So let's do some data access so our database get's created. You can run your application and you'll get your database Personal as specified in our config file. Add our first migration Migrations are a great resource as we can have a record for all the changes done and will generate the MySQL statements required to apply these changes to the database. Let's add a new property to our Person class public string Email { get; set; } If you try to run your application it will throw an exception saying  The model backing the 'PersonelContext' context has changed since the database was created. Consider using Code First Migrations to update the database (http://go.microsoft.com/fwlink/?LinkId=238269). So as suggested let's add our first migration for this change. In the Package Manager Console let's type Add-Migration AddEmailColumn Now we have the corresponding class which generate the necessary operations to update our database. namespace MigrationsFromScratch.Migrations { using System.Data.Entity.Migrations; public partial class AddEmailColumn : DbMigration { public override void Up(){ AddColumn("People", "Email", c => c.String(unicode: false)); } public override void Down() { DropColumn("People", "Email"); } } } In the Package Manager Console let's type Update-Database Now you can check your database to see all changes were succesfully applied. Now let's add a second change and generate our second migration public class Person   {       [Key]       public int PersonId { get; set;}       public string Name { get; set; }       public string Address {get; set;}       public string Email { get; set; }       public List<Skill> Skills { get; set; }   }   public class Skill   {     [Key]     public int SkillId { get; set; }     public string Description { get; set; }   }   public class PersonelContext : DbContext   {     public DbSet<Person> Persons { get; set; }     public DbSet<Skill> Skills { get; set; }   } If you would like to customize any part of this code you can do that at this step. You can see there is the up method which can update your database and the down that can revert the changes done. If you customize any code you should make sure to customize in both methods. Now let's apply this change. Update-database -verbose I added the verbose flag so you can see all the SQL generated statements to be run. Downgrading changes So far we have always upgraded to the latest migration, but there may be times when you want downgrade to a specific migration. Let's say we want to return to the status we have before our last migration. We can use the -TargetMigration option to specify the migration we'd like to return. Also you can use the -verbose flag. If you like to go  back to the Initial state you can do: Update-Database -TargetMigration:$InitialDatabase  or equivalent: Update-Database -TargetMigration:0  Migrations doesn't allow by default a migration that would ocurr in a data loss. One case when you can got this message is for example in a DropColumn operation. You can override this configuration by setting AutomaticMigrationDataLossAllowed to true in the configuration class. Also you can set your Database Initializer in case you want that these Migrations can be applied automatically and you don't have to go all the way through creating a migration and updating later the changes. Let's see how. Database Initialization by Code We can specify an initialization strategy by using Database.SetInitializer (http://msdn.microsoft.com/en-us/library/gg679461(v=vs.103)). One of the strategies that I found very useful when you are at a development stage (I mean not for production) is the MigrateDatabaseToLatestVersion. This strategy will make all the necessary migrations each time there is a change in our model that needs a database replication, this also implies that we have to enable AutomaticMigrationsEnabled flag in our Configuration class. public Configuration()         {             AutomaticMigrationsEnabled = true;             AutomaticMigrationDataLossAllowed = true;             SetSqlGenerator("MySql.Data.MySqlClient", new MySql.Data.Entity.MySqlMigrationSqlGenerator());    // This will add our MySQLClient as SQL Generator          } In the new EntityFramework section of your Config file we can set this at a context level basis.  The syntax is as follows: <contexts> <context type="Custom DbContext name, Assembly name"> <databaseInitializer type="System.Data.Entity.MigrateDatabaseToLatestVersion`2[[ Custom DbContext name, Assembly name],  [Configuration class name, Assembly name]],  EntityFramework" /> </context> </contexts> In our example this would be: The syntax is kind of odd but very convenient. This way all changes will always be applied when we do any data access in our application. There are a lot of new things to explore in EF 4.3.1 and Migrations so we'll continue writing some more posts about it. Please let us know if you have any questions or comments, also please check our forums here where we keep answering questions in general for the community.  Hope you found this information useful. Happy MySQL/.Net Coding! 

  Read the article

• NSFetchedResultsController crashing on performFetch: when using a cache

  - by Oliver

  I make use of NSFetchedResultsController to display a bunch of objects, which are sectioned using dates. On a fresh install, it all works perfectly and the objects are displayed in the table view. However, it seems that when the app is relaunched I get a crash. I specify a cache when initialising the NSFetchedResultsController, and when I don't it works perfectly. Here is how I create my NSFetchedResultsController: - (NSFetchedResultsController *)results { // If we are not nil, stop here if (results != nil) return results; // Create the fetch request, entity and sort descriptors NSFetchRequest *fetch = [[NSFetchRequest alloc] init]; NSEntityDescription *entity = [NSEntityDescription entityForName:@"Event" inManagedObjectContext:self.managedObjectContext]; NSSortDescriptor *descriptor = [[NSSortDescriptor alloc] initWithKey:@"utc_start" ascending:YES]; NSArray *descriptors = [[NSArray alloc] initWithObjects:descriptor, nil]; // Set properties on the fetch [fetch setEntity:entity]; [fetch setSortDescriptors:descriptors]; // Create a fresh fetched results controller NSFetchedResultsController *fetched = [[NSFetchedResultsController alloc] initWithFetchRequest:fetch managedObjectContext:self.managedObjectContext sectionNameKeyPath:@"day" cacheName:@"Events"]; fetched.delegate = self; self.results = fetched; // Release objects and return our controller [fetched release]; [fetch release]; [descriptor release]; [descriptors release]; return results; } These are the messages I get when the app crashes: FATAL ERROR: The persistent cache of section information does not match the current configuration. You have illegally mutated the NSFetchedResultsController's fetch request, its predicate, or its sort descriptor without either disabling caching or using +deleteCacheWithName: *** Terminating app due to uncaught exception 'NSInternalInconsistencyException', reason: 'FATAL ERROR: The persistent cache of section information does not match the current configuration. You have illegally mutated the NSFetchedResultsController's fetch request, its predicate, or its sort descriptor without either disabling caching or using +deleteCacheWithName:' I really have no clue as to why it's saying that, as I don't believe I'm doing anything special that would cause this. The only potential issue is the section header (day), which I construct like this when creating a new object: // Set the new format [formatter setDateFormat:@"dd MMMM"]; // Set the day of the event [event setValue:[formatter stringFromDate:[event valueForKey:@"utc_start"]] forKey:@"day"]; Like I mentioned, all of this works fine if there is no cache involved. Any help appreciated!

  Read the article

• ESB Toolkit.exceptionHandling Error - The application does not exist - Any ideas?

  - by Andrew Cripps

  Hello I am getting following error while attempting to run the Management Portal for ESB Toolkit 2.0: Event Type: Warning Event Source: ENTSSO Event Category: Enterprise Single Sign-On Event ID: 10536 Description: SSO AUDIT Function: GetConfigInfo (SSOProperties) Application Name: ESB Toolkit.exceptionHandling Error Code: 0xC0002A04, The application does not exist I am using SSO config store for the ESB Config. However, looking in the esb.config file, there is no section, like there is for the other esb SSO applications. Why might this section (and therefore the SSO app) be missing? How can I set it up? Thanks for any help with this Andrew

  Read the article

• Trouble Deploying .Net Framework 4 Website on IIS7

  - by Cyril Gupta

  Okay, I am trying to deploy a .Net framework 4 website on IIS7 server. I have already changed the application-pool's target framework to .Net 4, but the app is still showing me the error: "The configuration section 'system.web.extensions' cannot be read because it is missing a section declaration" I am guessing that has something to do with the new feature of .Net4 that lets me have a compact Web config file. I think for some reason IIS7 is not happy with this. What can I do to deploy this app successfully or do I have to scale back to v3.5? I am sure there is a solution out there. Do you have any suggestions?

  Read the article

• What is the number of the masked code ee@?

  - by Jack

  This question was asked in the interview of amdocs in the Unix section of the question paper.

  Read the article

• HTML5 cache manifest: whitelisting ALL remote resources?

  - by Alex Ford

  I'm doing an iPhone version of a desktop site that includes a blog. The blog often embeds images from other domains (the image URLs always start with http:// in this case, obviously), but because I'm using cache-manifest, these images don't load because they aren't declared in the manifest file. I have a NETWORK: whitelist section that has all of my AJAX request files, etc. I've even whitelisted the flickr farm domains because a lot of the images we add to the blog come from our flickr page. The flickr images show up just fine, but any other "random" image hotlinks from another domain show broken. I tried adding a line like this: http:// to the NETWORK: section, but it doesn't seem to like http:// as a whitelist. Does anyone have any thoughts on this? Thanks! Alex

  Read the article

• ASP.NET MVC Areas: How to hide "Area" name in URL?

  - by Mark Redman

  When running the MVC 2 Areas example that has a Blog Area and Blog Controller the URL looks like this: http://localhost:50526/Blog/Blog/ShowRecent in the format: RootUrl / AreaName / ControllerName / ActionName Having just discovered MVC Areas, it seem like a great way to organise code, ie create an Area for each section, which in my case each section has its own controller. This means that each AreaName = ControllerName. The effect of this is the double AreaName/ControllerName path in the Url eg /Blog/Blog/ above Not having a complete clear understanding of routing, how could I setup routing to not show the AreaName?

  Read the article

• Axis/SOAP service styles and interoperability

  - by Thilo

  There are four "styles" of service in Axis. RPC services use the SOAP RPC conventions, and also the SOAP "section 5" encoding. Document services do not use any encoding (so in particular, you won't see multiref object serialization or SOAP-style arrays on the wire) but DO still do XML<-Java databinding. Wrapped services are just like document services, except that rather than binding the entire SOAP body into one big structure, they "unwrap" it into individual parameters. Message services receive and return arbitrary XML in the SOAP Envelope without any type mapping / data binding. If you want to work with the raw XML of the incoming and outgoing SOAP Envelopes, write a message service. So, if I use anything else except the first option(SOAP RPC Section 5), how does this impact interoperability? If someone says they want a SOAP service (including WSDL), does this mean that SOAP RPC conventions are expected? Can the other three styles still be used when the other end is not implemented with Axis?

  Read the article

• IPv6 multicast addresses: Is the Group ID field effectively 112 bits or 32 bits?

  - by Jeremy Friesner

  Hi all, I'm trying to understand the rules for choosing an IPv6 multicast address Group ID, and the RFC seems somewhat inconsistent. For example, in RFC 2373 section 2.7 this diagram is shown: | 8 | 4 | 4 | 112 bits | +------ -+----+----+---------------------------------------------+ |11111111|flgs|scop| group ID | +--------+----+----+---------------------------------------------+ ... but then in section 2.7.2 it shows this: | 8 | 4 | 4 | 80 bits | 32 bits | +------ -+----+----+---------------------------+-----------------+ |11111111|flgs|scop| reserved must be zero | group ID | +--------+----+----+---------------------------+-----------------+ So my question is, are the upper 80 bits of the Group ID field usable or not? If they are usable, is it only under certain circumstances (e.g. when using non-Ethernet networking technology?) What problems should I expect to experience if I set these bits when multicasting over an Ethernet LAN?

  Read the article

• Convert XAML to FlowDocument to display in RichTextBox in WPF

  - by Erika

  I have some HTML, which i am converting to XAML using the library provided by Microsoft string t = HtmlToXamlConverter.ConvertHtmlToXaml(mail.HtmlDataString,true); now, from http://stackoverflow.com/questions/1449121/how-to-insert-xaml-into-richtextbox i am using the following: private static FlowDocument SetRTF(string xamlString) { StringReader stringReader = new StringReader(xamlString); System.Xml.XmlReader xmlReader = System.Xml.XmlReader.Create(stringReader); Section sec = XamlReader.Load(xmlReader) as Section; FlowDocument doc = new FlowDocument(); while (sec.Blocks.Count > 0) doc.Blocks.Add(sec.Blocks.FirstBlock); return doc; } This however keeps crashing unfortunately =/ Does anyone have any clue on how to display XAML text in a RichTextBox please?

  Read the article

• 301 htaccess redirect: add segment to old URL's

  - by Rick

  I'm trying to make sure old url's aren't broken after the site's URL structure has changed from this: http://www.domain.com/section/entry_name to this: http://www.domain.com/section/event_name/entry_name But to make it a bit more complex, I'm using the segments to sort entries, for example: http://www.domain.com/news/amazing_event/date/asc http://www.domain.com/videos/my_event/title/desc The new structure only effects one particular event (amazing_event) and should leave the other URL's alone. Where do I even begin to tackle this? My current .htaccess looks like: RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule ^(.*)$ /index.php/$1 [L] Thanks - appreciate any tips.

  Read the article

• FieldRenderer - Get specific field name from different template sections

  - by Zooking

  Hi, Is there a way to get a specific field name from a template with several sections to a FieldRenderer control? F.ex. I have a template with the sections "Data" and "Data2", both have a single-text-field called "Text". Is there a way to make my FieldRenderer get the field "Text" in section "Data2" It would be nice if one of the below suggestions worked: <sc:FieldRenderer ID="test" runat="server" FieldName="Text" Section="Data2" /> <sc:FieldRenderer ID="test" runat="server" FieldName="Data2/Text" /> BR Larre

  Read the article

• Problem in linking an nasm code

  - by Stefano

  I'm using a computer with an Intel Core 2 CPU and 2GB of RAM. The SO is Ubuntu 9.04. When I try to compile this code: ;programma per la simulazione di un terminale su PC, ottenuto utilizzando l'8250 ;in condizione di loopback , cioè Tx=Rx section .code64 section .data TXDATA EQU 03F8H ;TRASMETTITORE RXDATA EQU 03F8H ;RICEVITORE BAUDLSB EQU 03F8H ;DIVISORE DI BAUD RATE IN LSB BAUDMSB EQU 03F9H ;DIVISORE DI BAUD RATE IN MSB INTENABLE EQU 03F9H ;REGISTRO DI ABILITAZIONE DELL'INTERRUZIONE INTIDENTIF EQU 03FAH ;REGISTRO DI IDENTIFICAZIONE DELL'INTERRUZIONE LINECTRL EQU 03FBH ;REGISTRO DI CONTROLLO DELLA LINEA MODEMCTRL EQU 03FCH ;REGISTRO DI CONTROLLO DEL MODEM LINESTATUS EQU 03FDH ;REGISTRO DI STATO DELLA LINEA MODEMSTATUS EQU 03FEH ;REGISTRO DI STATO DEL MODEM BAUDRATEDIV DW 0060H ;DIVISOR: LOW=60, HIGH=00 -BAUD =9600 COUNTERCHAR DB 0 ;CHARACTER COUNTER ;DW 256 DUP (?) section .text global _start _start: ;PROGRAMMAZIONE 8250 MOV DX,LINECTRL MOV AL,80H ;BIT 7=1 PER INDIRIZZARE IL BAUD RATE OUT DX,AL MOV DX,BAUDLSB MOV AX,BAUDRATEDIV ;DEFINISCO FATTORE DI DIVISIONE OUT DX,AL MOV DX,BAUDMSB MOV AL,AH OUT DX,AL ;MSB MOV DX,LINECTRL MOV AL,00000011B ;8 BIT DATO, 1 STOP, PARITA' NO OUT DX,AL MOV DX,MODEMCTRL MOV AL,00010011B ;BIT 4=0 PER NO LOOPBACK OUT DX,AL MOV DX,INTENABLE XOR AL,AL ;DISABILITO TUTTI GLI INTERRUPTS OUT DX,AL CICLO: MOV DX,LINESTATUS IN AL,DX ;LEGGO IL REGISTRO DI STATO DELLA LINEA TEST AL,00011110B ;VERIFICO GLI ERRORI (4 TIPI) JNE ERRORI TEST AL,01H ;VERIFICO Rx PRONTO JNE LEGGOCHAR TEST AL,20H ;VERIFICO Tx VUOTO JE CICLO ;SE SI ARRIVA A QUESTO PUNTO ALLORA L'8250 è PRONTO PER TRASMETTERE UN NUOVO CARATTERE MOV AH,1 INT 80H JE CICLO ;SE SI ARRIVA A QUESTO PUNTO SIGNIFICA CHE ESISTE UN CARATTERE DA TASTIERA MOV AH,0 INT 80H ;Al CONTIENE IL CARATTERE DELLA TASTIERA MOV DX,3F8H OUT DX,AL JMP CICLO LEGGOCHAR: MOV AL,[COUNTERCHAR] INC AL CMP AL,15 JE FINE MOV [COUNTERCHAR],AL MOV DX,TXDATA IN AL,DX ;AL CONTIENE IL CARATTERE RICEVUTO AND AL,7FH ;POICHè VI SONO 7 BIT DI DATO ;VISUALIZZAZIONE DEL CARATTERE MOV BX,0 MOV AH,14 INT 80H POP AX CMP AL,0DH ;CONTROLLO SE RETURN JNE CICLO ;CAMBIO RIGA DI VISUALIZZAZIONE MOV AL,0AH MOV BX,0 MOV AH,14 ;INT 10H INT 80H JMP CICLO ;GESTIONE ERRORI ERRORI: MOV DX,3F8H IN AL,DX MOV AL,'?' MOV BX,0 MOV AH,14 INT 80H JMP CICLO FINE: XOR AH,AH MOV AL,03 INT 80H When I compile this code "NASM -f bin UARTLOOP.asm", the compiler can create the UARTLOOP.o file without any error. When I try to link the .o file with "ld UARTLOOP.o" it tells: UARTLOOP.o: In function `_start': UARTLOOP.asm:(.text+0xd): relocation truncated to fit: R_X86_64_16 against `.data' Have u got some ideas to solve this problem? Thx =)

  Read the article

• Nice name for `decorator' class?

  - by Lajos Nagy

  I would like to separate the API I'm working on into two sections: 'bare-bones' and 'cushy'. The idea is that all method calls in the 'cushy' section could be expressed in terms of the ones in the 'bare-bones' section, that is, they would only serve as convenience methods for the quick-and-dirty. The reason I would like to do this is that very often when people are beginning to use an API for the first time, they are not interested in details and performance: they just want to get it working. Anybody tried anything similar before? I'm particularly interested in naming conventions and organizing the code.

  Read the article

• web.config, configSource, and "The 'xxx' element is not declared" warning.

  - by UpTheCreek

  I have broken down the horribly unwieldy web.config file into individual files for some of the sections (e.g. connectionStrings, authentication, pages etc.) using the configSource attribute. This is working file, but the individual xml files that hold the section 'snippets' cause warnings in VS. For example, a file named roleManager.config is used for the role manager section, and looks like this: <roleManager enabled="false"> </rolemanager> However I get a blue squiggle under the roleManager element in VS, and the following warning: The 'roleManager' element is not declared I guess this is something to do with valid xml and schemas etc. Is there an easy way to fix this? Something I can add to the individual files? Thanks P.S. I have heard that it is bad practice to break the web.config file out like this. But don't really understand why - can anyone illuminate me?

  Read the article

• GD DLL installation

  - by smokinguns

  I'm using the JpGraph library(PHP graphing Lib). I'm getting the foll error: "JpGraph Error Your PHP installation does not seem to have the required GD 2.x library enabled. Please see the PHP documentation, "Image" section. Make sure that "php_gd2.dll" statement is uncomment in the [modules] section in the php.ini file" I uncommented the php_gd2.dll statement in php.ini(in C:\WINDOWS) and set the 'extension_dir' as extension_dir = C:\PHP\extensions (I have a extensions directory and put the php_gd2.dll in this directory ). I restarted IIS and when I fire up my webapp, it doesnt comeup. I commented out the php_gd2.dll and restarted IIS again. The webapp came up. I dont understand what I'm doing wrong. Here is the server config: Windows Server 2003, with IIS Version 6. The PHP version is 4.3.8

  Read the article

• nested for loop

  - by Gary

  Hello, Just learning Python and trying to do a nested for loop. What I'd like to do in the end is place a bunch of email addresses in a file and have this script find the info, like the sending IP of mail ID. For now i'm testing it on my /var/log/auth.log file Here is my code so far: #!/usr/bin/python # this section puts emails from file(SpamEmail) in to a array(array) in_file = open("testFile", "r") array = in_file.readlines() in_file.close() # this section opens and reads the target file, in this case 'auth.log' log = open("/var/log/auth.log", "r") auth = log.readlines() for email in array: print "Searching for " +email, for line in auth: if line.find(email) > -1: about = line.split() print about[0], print Inside 'testfile' I have the word 'disconnect' cause I know it's in the auth.log file. It just doesn't find the word 'disconnect'. In the line of "if line.find(email) -1:" i can replace email and put "disconnect" the scripts finds it fine. Any idea? Thanks in advance. Gary

  Read the article

• Set text highlight colour of Blackberry RichTextField

  - by DaveJohnston

  Does anyone know how to set the background colour of just a section of text within a RichTextField on the Blackberry? I already use the offsets, attributes and fonts arrays to make changes to the appearance of certain sections of the text, but I would like to add a highlight colour to the background of one section too. I know there is a protected method called getBackgroundColors that returns an array of colors to be used, which I can overwrite. But I have tried this and that method never seems to get called in my code, I don't actually know how and when the underlying implementation of the RichTextField actually use this method. Any ideas?

  Read the article

• What is the semantically correct way to use the `<article>` tag in HTML 5, with `<ol>, <ul>, and <li

  - by viatropos

  I currently have an ordered list that I want to markup using the new HTML 5 attributes. It looks like this: <ol class="section"> <li class="article"> <h2>Article A</h2> <p>Some text</p> </li> <li class="article"> <h2>Article B</h2> <p>Some text</p> </li> <li class="article"> <h2>Article C</h2> <p>Some text</p> </li> </ol> It seems the only way to keep the list AND use HTML 5 tags is to add a whole bunch of unnecessary divs: <section> <ol> <li> <article> <h2>Article A</h2> <p>Some text</p> </article> </li> <li> <article> <h2>Article B</h2> <p>Some text</p> </article> </li> <li> <article> <h2>Article C</h2> <p>Some text</p> </article> </li> </ol> </section> Is there a better way to do this? What are your thoughts?

  Read the article

• AJAX in Drupal Forms?

  - by Kevin

  How would you go about constructing a step by step form that uses AJAX through Drupal to pull the next form step? For example, Step 1: I like Baseball I don't like Baseball. When that person clicks on either Like or Don't Like, I want to use AJAX to recognize and pull the next part of the form, remove/hide the first section since its not needed, and present the next section. Example: Step 1: I like Baseball *click (fade out) Step 2: My favorite team is __________ The player I like most is __________ What is the best way to do this through Drupal Form API? I know how to build the forms and modules, but I have never used AJAX yet. I know a few things exist out there that are supposed to help, but I wanted to know if anyone here has done it and how they approached it.

  Read the article

• Can NSDictionary be used with TableView on iPhone?

  - by bobo

  In a UITableViewController subclass, there are some methods that need to be implemented in order to load the data and handle the row selection event: - (NSInteger)numberOfSectionsInTableView:(UITableView *)tableView { return 1; //there is only one section needed for my table view } - (NSInteger)tableView:(UITableView *)tableView numberOfRowsInSection:(NSInteger)section { return [myList count]; //myList is a NSDictionary already populated in viewDidLoad method } - (UITableViewCell *)tableView:(UITableView *)tableView cellForRowAtIndexPath:(NSIndexPath *)indexPath { static NSString *CellIdentifier = @"Cell"; UITableViewCell *cell = [tableView dequeueReusableCellWithIdentifier:CellIdentifier]; if (cell == nil) { cell = [[[UITableViewCell alloc] initWithStyle:UITableViewCellStyleDefault reuseIdentifier:CellIdentifier] autorelease ]; } // indexPath.row returns an integer index, // but myList uses keys that are not integer, // I don't know how I can retrieve the value and assign it to the cell.textLabel.text return cell; } - (void)tableView:(UITableView *)tableView didSelectRowAtIndexPath:(NSIndexPath *)indexPath { // Handle row on select event, // but indexPath.row only returns the index, // not a key of the myList NSDictionary, // this prevents me from knowing which row is selected } How is NSDictionary supposed to work with TableView? What is the simplest way to get this done?

  Read the article

< Previous Page | 44 45 46 47 48 49 50 51 52 53 54 55  | Next Page >