Search Results

Search found 2993 results on 120 pages for 'distributed transactions'.

Page 93/120 | < Previous Page | 89 90 91 92 93 94 95 96 97 98 99 100  | Next Page >

• How to manage processes-to-CPU cores affinities ?

  - by Philippe

  I use a distributed user-space filesystem (GlusterFS) and I would like to be sure GlusterFS processes will always have the computing power they need. Each execution node of my grid have 2 CPU, with 4 cores per CPU and 2 threads per core (16 "processors" are seen by Linux). My goal is to guarantee that GlusterFS processes have enough processing power to be reliable, responsive and fast. (There is no marketing here, just the dreams of a sysadmin ;-) I consider two main points : GlusterFS processes I/O for data access (on local disks, or remote disks) I thought about binding the Linux Kernel and GlusterFS instances on a specific "processor". I would like to be sure that : No grid job will impact the kernel and the GlusterFS instances Researchers jobs won't be affected by system processes (I'd like to reserve a pool of cores to job execution and be sure that no system process will use these CPUs) But what about I/O ? As we handle a huge amount of data (several terabytes), we'll have a lot of interuptions. How can I distribute these operations on my processors ? What are the "best practices" ? Thanks for your comments!

  Read the article

• Windows Azure Recipe: Consumer Portal

  - by Clint Edmonson

  Nearly every company on the internet has a web presence. Many are merely using theirs for informational purposes. More sophisticated portals allow customers to register their contact information and provide some level of interaction or customer support. But as our understanding of how consumers use the web increases, the more progressive companies are taking advantage of social web and rich media delivery to connect at a deeper level with the consumers of their goods and services. Drivers Cost reduction Scalability Global distribution Time to market Solution Here’s a sketch of how a Windows Azure Consumer Portal might be built out: Ingredients Web Role – this will host the core of the solution. Each web role is a virtual machine hosting an application written in ASP.NET (or optionally php, or node.js). The number of web roles can be scaled up or down as needed to handle peak and non-peak traffic loads. Database – every modern web application needs to store data. SQL Azure databases look and act exactly like their on-premise siblings but are fault tolerant and have data redundancy built in. Access Control (optional) – if identity needs to be tracked within the solution, the access control service combined with the Windows Identity Foundation framework provides out-of-the-box support for several social media platforms including Windows LiveID, Google, Yahoo!, Facebook. It also has a provider model to allow integration with other platforms as well. Caching (optional) – for sites with high traffic with lots of read-only data and lists, the distributed in-memory caching service can be used to cache and serve up static data at higher scale and speed than direct database requests. It can also be used to manage user session state. Blob Storage (optional) – for sites that serve up unstructured data such as documents, video, audio, device drivers, and more. The data is highly available and stored redundantly across data centers. Each entry in blob storage is provided with it’s own unique URL for direct access by the browser. Content Delivery Network (CDN) (optional) – for sites that service users around the globe, the CDN is an extension to blob storage that, when enabled, will automatically cache frequently accessed blobs and static site content at edge data centers around the world. The data can be delivered statically or streamed in the case of rich media content. Training Labs These links point to online Windows Azure training labs where you can learn more about the individual ingredients described above. (Note: The entire Windows Azure Training Kit can also be downloaded for offline use.) Windows Azure (16 labs) Windows Azure is an internet-scale cloud computing and services platform hosted in Microsoft data centers, which provides an operating system and a set of developer services which can be used individually or together. It gives developers the choice to build web applications; applications running on connected devices, PCs, or servers; or hybrid solutions offering the best of both worlds. New or enhanced applications can be built using existing skills with the Visual Studio development environment and the .NET Framework. With its standards-based and interoperable approach, the services platform supports multiple internet protocols, including HTTP, REST, SOAP, and plain XML SQL Azure (7 labs) Microsoft SQL Azure delivers on the Microsoft Data Platform vision of extending the SQL Server capabilities to the cloud as web-based services, enabling you to store structured, semi-structured, and unstructured data. Windows Azure Services (9 labs) As applications collaborate across organizational boundaries, ensuring secure transactions across disparate security domains is crucial but difficult to implement. Windows Azure Services provides hosted authentication and access control using powerful, secure, standards-based infrastructure. See my Windows Azure Resource Guide for more guidance on how to get started, including links web portals, training kits, samples, and blogs related to Windows Azure.

  Read the article

• "Delivered-To" Header in Exchange

  - by Kaii

  In some SMTP server implementations (i.e. Postfix) you can enable Delivered-To and X-Original-To headers that will be added to your email. (or [X-]Envelope-To) This is very helpful with distribution lists to determine which e-mail address the mail has been redirected to. So, when the mail has been sent to [email protected], you can see in the Delivered-To or Envelope-To header that it has been redirected (distributed) to [email protected], which is one of many other e-mail addresses that are linked to a single mailbox. How do I find which address was used to deliver this mail to a specific mailbox on Microsoft Exchange 2010? Looking at the plain message (with all headers) i can not find any information that the mail arrived via address [email protected] I think I need the Delivered-To header (or a similar one) to be set on Microsoft Exchange when a mail is delivered via distribution lists. Is there any way to enable such header in Exchange 2010? I need it so that our Ticket system (OTRS) correctly recognizes where the ticket belongs to. Adding all the e-mail addresses of all distribution lists to the system configuration is not the right solution. And if there is a solution for Exchange 2010, is this possibly also applicable to Exchange 2007?

  Read the article

• Wisdom Lies in Collaborative Power and Intelligence

  - by kellsey.ruppel

  By Alakh Verma, Director, Platform Technology Solutions   In my recent blog posts, I shared insights on Predictive Analytics (Will Predictive Analytics at 'Speed of Thoughts' Help Businesses?), Real Time Decisions (How critical are Real Time decisions in business today?) and their significance in our lives in general and in businesses today. In the current business paradigm shift- with evolutionary social business, it is paramount that businesses look for wisdom in collaborative power and intelligence and equip their employees with the tools to engage with one another. There is an old time saying that 5 sticks tied together are stronger and unable to break as opposed to an individual stick. We have recently witnessed the power of ordinary people uniting together and fought collaboratively using Facebook and Twitter to topple down dictators in Tunisia, Egypt and Libya—and are threatening absolute rule in Syria. And an India one man’s (Anna Hazare) campaign against corruption went viral, bringing thousands to the streets in support. As anyone who has worked in a sizeable organization knows, there is no guarantee that the organization as a whole will perform efficiently and achieve its goals, even if each employee is individually efficient and every team has a high level of productivity. To achieve enterprise productivity, it is necessary not only for individuals and groups to “do things right” by working productively but also for the enterprise as a whole to “do the right things” - form the right teams, make the right decisions, allocate resources correctly, and effectively coordinate activities across the entire organization. Most organizations fall short of the optimal level of enterprise productivity because of one or more of these reasons, all at a great cost to the business.  They are disconnected from themselves with various parts of the organization unintentionally working at cross-purposes with each other.  Information that exists is not getting shared or reused.  Human talent is not being applied where it is most needed.  The same problems are being solved repeatedly by multiple groups. Intelligent collaboration through automated business processes has the ability to alter the course of any important business activity, with a potentially dramatic impact on the financial performance of the business. Whether it is a simple email exchange, a physical or virtual meeting, a task force, or a large-scale project, the activity is inherently collaborative.  In fact, collaboration can be defined as the work that takes place among people when a business process is not pre-determining how the work should take place. Collaboration is many things: information sharing, brainstorming, problem solving, best practice negotiation, innovation, coordination of activity, alignment of purpose, and so forth.  Collaboration is the “white space” between the business processes; it is the glue that holds an organization together, and the lubricant that allows the machinery to keep running.  Real time search and collaborative capabilities of the right people with the right content supported by defined processes will provide unparallel wisdom in the organization in the most competitive business environment today. Interestingly, technologies such as Oracle WebCenter offer these capabilities in our Web based business transactions and compliment in the overall collaborative intelligence and power to truly transform organizations to social businesses. Looking to learn more about engaging your employees to collaborate together and providing a complete user experience for your customers? You won't want to miss our webcast today! Drive Online Engagement with Intuitive Portals and Websites

  Read the article

• SQL SERVER – Last Two Days to Get FREE Book – Joes 2 Pros Certification 70-433

  - by pinaldave

  Earlier this week we announced that we will be giving away FREE SQL Wait Stats book to everybody who will get SQL Server Joes 2 Pros Combo Kit. We had a fantastic response to the contest. We got an overwhelming response to the offer. We knew there would be a great response but we want to honestly say thank you to all of you for making it happen. Rick and I want to make sure that we express our special thanks to all of you who are reading our books. The offer is still on and there are two more days to avail this offer. We want to make sure that everybody who buys our most selling combo kits, we will send our other most popular SQL Wait Stats book. Please read all the details of the offer here. The books are great resources for anyone who wants to learn SQL Server from fundamentals and eventually go on the certification path of 70-433. Exam 70-433 contains following important subject and the book covers the subject of fundamental. If you are taking the exam or not taking the exam – this book is for every SQL Developer to learn the subject from fundamentals.  Create and alter tables. Create and alter views. Create and alter indexes. Create and modify constraints. Implement data types. Implement partitioning solutions. Create and alter stored procedures. Create and alter user-defined functions (UDFs). Create and alter DML triggers. Create and alter DDL triggers. Create and deploy CLR-based objects. Implement error handling. Manage transactions. Query data by using SELECT statements. Modify data by using INSERT, UPDATE, and DELETE statements. Return data by using the OUTPUT clause. Modify data by using MERGE statements. Implement aggregate queries. Combine datasets. INTERSECT, EXCEPT Implement subqueries. Implement CTE (common table expression) queries. Apply ranking functions. Control execution plans. Manage international considerations. Integrate Database Mail. Implement full-text search. Implement scripts by using Windows PowerShell and SQL Server Management Objects (SMOs). Implement Service Broker solutions. Track data changes. Data capture Retrieve relational data as XML. Transform XML data into relational data. Manage XML data. Capture execution plans. Collect output from the Database Engine Tuning Advisor. Collect information from system metadata. Availability of Book USA - Amazon | India - Flipkart | Indiaplaza Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Joes 2 Pros, PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

  Read the article

• Comparing Isis, Google, and Paypal

  - by David Dorf

  Back in 2010 I was sure NFC would make great strides, but here we are two years later and NFC doesn't seem to be sticking. The obvious reason being the chicken-and-egg problem.  Retailers don't want to install the terminals until the phones support NFC, and vice-versa. So consumers continue to sit on the sidelines waiting for either side to blink and make the necessary investment.  In the meantime, EMV is looking for a way to sneak into the US with the help of the card brands. There are currently three major solutions that are battling in the marketplace.  All three know that replacing mag-stripe alone is not sufficient to move consumers.  Long-term it's the offers and loyalty programs combined with tendering that make NFC attractive. NFC solutions cross lots of barriers, so a strong partner system is required.  The solutions need to include the carriers, card brands, banks, handset manufacturers, POS terminals, and most of all lots of merchants.  Lots of coordination is necessary to make the solution seamless to the consumer. Google Wallet Google's problem has always been that only the Nexus phone has an NFC chip that supports their wallet.  There are a couple of additional phones out there now, but adoption is still slow.  They acquired Zavers a while back to incorporate digital coupons, but the the bulk of their users continue to be non-NFC.  They have taken an open approach by not specifying particular payment brands.  Google is piloting in San Francisco and New York, supporting both MasterCard PayPass and stored value. I suppose the other card brands may eventually follow.  There's no cost for consumers or merchants -- Google will make money via targeted ads. Isis Not long after Google announced its wallet, AT&T, Verizon, and T-Mobile announced a joint venture called Isis.  They are in the unique position of owning the SIM in the phones they issue.  At first it seemed Isis was a vehicle for the carriers to compete with the existing card brands, but Isis later switched to a generic wallet that supports the major card brands.  Isis reportedly charges issuers a $5 fee per customer per year.  Isis will pilot this summer in Salt Lake City and Austin. PayPal PayPal, the clear winner in the online payment space beyond traditional credit cards, is trying to move into physical stores.  After negotiations with Google to provide a wallet broke off, PayPal decided to avoid NFC altogether, at least for now, and focus on payments without any physical card or phone.  By avoiding NFC, consumers don't need an NFC-enabled phone and merchants don't need a new reader.  Consumers must enter their phone number and PIN in the merchant's existing device, or they can enter their PIN in the PayPal inStore app running on their phone, then show the merchant a unique barcode which authorizes payment. Paypal is free for consumers and charges a fee for merchants.  Its not clear, at least to me, how PayPal handles fraudulent transactions and whether the consumer is protected. The wildcard is, of course, Apple.  Their mobile technologies set the standard, so incorporating NFC chips would certainly accelerate adoption of many payment solutions.  Their announcement today of the iOS Passbook is a step in the right direction, but stops short of handling payments. For those retailers that have invested in modern terminals, it seems the best strategy is to support all the emerging solutions and let the consumers choose the winner.

  Read the article

• SQL Server IO handling mechanism can be severely affected by high CPU usage

  - by sqlworkshops

  Are you using SSD or SAN / NAS based storage solution and sporadically observe SQL Server experiencing high IO wait times or from time to time your DAS / HDD becomes very slow according to SQL Server statistics? Read on… I need your help to up vote my connect item – https://connect.microsoft.com/SQLServer/feedback/details/744650/sql-server-io-handling-mechanism-can-be-severely-affected-by-high-cpu-usage. Instead of taking few seconds, queries could take minutes/hours to complete when CPU is busy.In SQL Server when a query / request needs to read data that is not in data cache or when the request has to write to disk, like transaction log records, the request / task will queue up the IO operation and wait for it to complete (task in suspended state, this wait time is the resource wait time). When the IO operation is complete, the task will be queued to run on the CPU. If the CPU is busy executing other tasks, this task will wait (task in runnable state) until other tasks in the queue either complete or get suspended due to waits or exhaust their quantum of 4ms (this is the signal wait time, which along with resource wait time will increase the overall wait time). When the CPU becomes free, the task will finally be run on the CPU (task in running state).The signal wait time can be up to 4ms per runnable task, this is by design. So if a CPU has 5 runnable tasks in the queue, then this query after the resource becomes available might wait up to a maximum of 5 X 4ms = 20ms in the runnable state (normally less as other tasks might not use the full quantum).In case the CPU usage is high, let’s say many CPU intensive queries are running on the instance, there is a possibility that the IO operations that are completed at the Hardware and Operating System level are not yet processed by SQL Server, keeping the task in the resource wait state for longer than necessary. In case of an SSD, the IO operation might even complete in less than a millisecond, but it might take SQL Server 100s of milliseconds, for instance, to process the completed IO operation. For example, let’s say you have a user inserting 500 rows in individual transactions. When the transaction log is on an SSD or battery backed up controller that has write cache enabled, all of these inserts will complete in 100 to 200ms. With a CPU intensive parallel query executing across all CPU cores, the same inserts might take minutes to complete. WRITELOG wait time will be very high in this case (both under sys.dm_io_virtual_file_stats and sys.dm_os_wait_stats). In addition you will notice a large number of WAITELOG waits since log records are written by LOG WRITER and hence very high signal_wait_time_ms leading to more query delays. However, Performance Monitor Counter, PhysicalDisk, Avg. Disk sec/Write will report very low latency times.Such delayed IO handling also occurs to read operations with artificially very high PAGEIOLATCH_SH wait time (with number of PAGEIOLATCH_SH waits remaining the same). This problem will manifest more and more as customers start using SSD based storage for SQL Server, since they drive the CPU usage to the limits with faster IOs. We have a few workarounds for specific scenarios, but we think Microsoft should resolve this issue at the product level. We have a connect item open – https://connect.microsoft.com/SQLServer/feedback/details/744650/sql-server-io-handling-mechanism-can-be-severely-affected-by-high-cpu-usage - (with example scripts) to reproduce this behavior, please up vote the item so the issue will be addressed by the SQL Server product team soon.Thanks for your help and best regards,Ramesh MeyyappanHome: www.sqlworkshops.comLinkedIn: http://at.linkedin.com/in/rmeyyappan

  Read the article

• Redis as substitution for Memcache

  - by Boban P.

  We have distributed web app, and for now, as session handler, we use two separate instances of memcache in redundancy, so everything that is written in one memcache is also written in other. Memcache is fairly easy to install, use, and maintain but we have one problem: if one memcache fail, everything is fine, php comunicate with other instance which has all data (although, half of connections have a delay because they try to use failed one, wait a little, and then contact other memcache). When failed instance comes back to life again, it starts up empty. If established session request data from that instance, session fails, and user logs out, and that happens to half of users.So, we are thinking about to switch to redis for session handling, and maybe keep memcache for cache only. My questions are: If we setup redis instances as master-slave, and if master fails, can sentinel promote slave as new master and when old master comes back to life, will it stay as slave or not? Is redis call malloc at startup to allocate part of memory, like memcache or varnish, or it calls malloc for every key inserted? And what are pros and cons of that?

  Read the article

• What is the difference between Workcenters, Dashboards, and the Interaction Hub?

  - by Matthew Haavisto

  Oracle Open World has just concluded.  Over the course of the conference, we presented several sessions covering different aspects of the PeopleSoft user experience, including Workcenters, Dashboards, and the PeopleSoft Interaction Hub (formerly known as the PeopleSoft Applications Portal).  Although we've produced collateral on these features and covered them in sessions, it became apparent at the conference that customers still have many questions about the these products, including how they are licensed, how they are installed, what their various purposes are, and how they can be used together synergistically. Let's Start with Licensing and Installation As you may know, we've extended the restricted use license (RUL) for the Interaction Hub.  This grants customers with PeopleTools 8.52 licenses the right to install the Interaction Hub for free for use as specified in the Tools license notes.  Note that this means customers receive a restricted use license for the Interaction Hub that doesn't cost them an additional license fee, but it is a separate product, not part of PeopleTools or PeopleSoft applications, and is a separate installation.  This means customers must provide the infrastructure to install and run the Hub, just like any other application.  The benefits of using the Hub to unify your PeopleSoft user experience can be great.  PeopleSoft applications have not yet delivered instances of the Hub with their products, though they may in the future. Workcenters and Dashboards, on the other hand, are frameworks provided by PeopleTools.  No other license is required, and no additional installation of a separate product is needed (apart from PeopleTools and PeopleSoft applications).  PeopleSoft applications are delivering instances of the workcenters and dashboards with their products.  Some are available now, and more are coming in future releases.  These delivered workcenter and dashboard instances require no additional licenses, and no additional installations beyond Tools and the applications that provide them.  In addition, the workcenter and dashboard frameworks provided by PeopleTools can be used by customers to build their own workcenters and dashboards, and it's quite easy and simple to do so. What are Their Differences?  What Purposes do they Serve? Workcenters, Dashboards and the Interaction Hub appear somewhat similar.  They all contain pagelets, and have some visual characteristics in common.  However, their strengths and purposes are very different, and they were designed to provide different benefits to your PeopleSoft ecosystem. Workcenters and Dashboards have the following characteristics: Designed for specific roles Focus on the daily tasks of those roles Help to streamline the work performed most often Personal view of my work world Makes navigation and search easier and quicker, particularly for transactions and decision support Reports and data needed for day-to-day work Personalizable, but minimal Delivered by PS Apps, but can be altered by customer for their requirements Customers can create their own Workcenters can be used for guided processes  The Interaction Hub is designed to aggregate content from multiple applications, and is is used to unify the user experience of those applications.  It offers a rich, web site-based user experience, and is often used to provide access to infrequently performed activities like benefits enrollment, payroll inquiries, life event changes, onboarding, and so on. Full-featured and robust Centrally administered Pushed to large audience Broad info like Company News Infrequent activities like benefits, not day-to-day tasks Self-service, access to employer info Central launch point for other activities and can navigate to workcenters and dashboards Deployed by customers or consultants, instances not delivered by PeopleSoft (at this time) Content management Unified PS application navigation Although these products are quite different and serve different purposes in your PeopleSoft environment, they can be used together to provide a richer, more efficient and engaging user experience for your all your user communities.

  Read the article

• Understanding the Customer Form in Release 12 from an AR Perspective!!

  - by user793553

  Confused by the Customer Form in Release 12??  Read on, to get some insight into the evolution of this screen, and how it links in with Trading Community Architecture. Historically, the customer data model was owned by Oracle Receivables (AR).  However, as the data model changed and more complex relationships and attributes had to be tracked and monitored, the Trading Community Architecture (TCA) product was created.  All applications within the E-Business suite that require interaction with a customer integrate with TCA. Customer information is no longer stored in the individual applications but rather in a central repository/registry maintained within TCA.  It is important to understand the following entities/concepts stored in TCA: Party: A party is an entity with whom you can have a potential business relationship.  A party can be either a Person or an Organization.  The Party entity is completely independent of any business relationship; this means that a Party can exist even if you have no transactions with it.   The Party is the "umbrella" entity under which you capture all other attributes listed below. Customer: A customer is a party with whom you have an existing business relationship.  From an AR perspective, you can simplify the concepts by thinking of a Customer as a Party. This definition however does not apply to all other applications. In the Oracle Receivables Customer form, the information displayed at the Customer level is from TCA's Party information record. Customer Account (also called Account): An account contains information about how you transact business with a particular customer.  You can create multiple accounts for a customer.  When you create invoices and receipts you associate it to a particular Account of a Customer. Location: A Location is an address.  It is a point in space, typically identified by a street number, a street name, a city, a state or province, a country.  A location is independent of what it is used for - you do not associate a purpose to a location. Party Site: A Party Site is associated to a Party.  It is the location where a party is physically located.  When defining sites for a Party, only one can be an identifying address.  However, you can define other party sites associated to a party. You can define purposes/usage for Party Sites. Account Site: An Account Site is associated to a Customer Account. It is the location associated to the account you are transacting business with. You can define business purposes (also called site uses) for an Account site. Read more about the Customer Workbench in these notes: Doc ID 1436547.1 Oracle Receivables: Understanding the Customer Form in Release 12 Doc ID  1437866.1 Customer Form - Address: Troubleshooting, Known Issues and Patches Doc ID  1448442.1 Oracle Receivables (AR): Customer Workbench Information Center Do you find this type of blog entry useful?  Please add comments to let us know how we can help you more effectively.  Thank you!

  Read the article

• Oracle MAA Part 1: When One Size Does Not Fit All

  - by JoeMeeks

  The good news is that Oracle Maximum Availability Architecture (MAA) best practices combined with Oracle Database 12c (see video) introduce first-in-the-industry database capabilities that truly make unplanned outages and planned maintenance transparent to users. The trouble with such good news is that Oracle’s enthusiasm in evangelizing its latest innovations may leave some to wonder if we’ve lost sight of the fact that not all database applications are created equal. Afterall, many databases don’t have the business requirements for high availability and data protection that require all of Oracle’s ‘stuff’. For many real world applications, a controlled amount of downtime and/or data loss is OK if it saves money and effort. Well, not to worry. Oracle knows that enterprises need solutions that address the full continuum of requirements for data protection and availability. Oracle MAA accomplishes this by defining four HA service level tiers: BRONZE, SILVER, GOLD and PLATINUM. The figure below shows the progression in service levels provided by each tier. Each tier uses a different MAA reference architecture to deploy the optimal set of Oracle HA capabilities that reliably achieve a given service level (SLA) at the lowest cost.  Each tier includes all of the capabilities of the previous tier and builds upon the architecture to handle an expanded fault domain. Bronze is appropriate for databases where simple restart or restore from backup is ‘HA enough’. Bronze is based upon a single instance Oracle Database with MAA best practices that use the many capabilities for data protection and HA included with every Oracle Enterprise Edition license. Oracle-optimized backups using Oracle Recovery Manager (RMAN) provide data protection and are used to restore availability should an outage prevent the database from being able to restart. Silver provides an additional level of HA for databases that require minimal or zero downtime in the event of database instance or server failure as well as many types of planned maintenance. Silver adds clustering technology - either Oracle RAC or RAC One Node. RMAN provides database-optimized backups to protect data and restore availability should an outage prevent the cluster from being able to restart. Gold raises the game substantially for business critical applications that can’t accept vulnerability to single points-of-failure. Gold adds database-aware replication technologies, Active Data Guard and Oracle GoldenGate, which synchronize one or more replicas of the production database to provide real time data protection and availability. Database-aware replication greatly increases HA and data protection beyond what is possible with storage replication technologies. It also reduces cost while improving return on investment by actively utilizing all replicas at all times. Platinum introduces all of the sexy new Oracle Database 12c capabilities that Oracle staff will gush over with great enthusiasm. These capabilities include Application Continuity for reliable replay of in-flight transactions that masks outages from users; Active Data Guard Far Sync for zero data loss protection at any distance; new Oracle GoldenGate enhancements for zero downtime upgrades and migrations; and Global Data Services for automated service management and workload balancing in replicated database environments. Each of these technologies requires additional effort to implement. But they deliver substantial value for your most critical applications where downtime and data loss are not an option. The MAA reference architectures are inherently designed to address conflicting realities. On one hand, not every application has the same objectives for availability and data protection – the Not One Size Fits All title of this blog post. On the other hand, standard infrastructure is an operational requirement and a business necessity in order to reduce complexity and cost. MAA reference architectures address both realities by providing a standard infrastructure optimized for Oracle Database that enables you to dial-in the level of HA appropriate for different service level requirements. This makes it simple to move a database from one HA tier to the next should business requirements change, or from one hardware platform to another – whether it’s your favorite non-Oracle vendor or an Oracle Engineered System. Please stay tuned for additional blog posts in this series that dive into the details of each MAA reference architecture. Meanwhile, more information on Oracle HA solutions and the Maximum Availability Architecture can be found at: Oracle Maximum Availability Architecture - Webcast Maximize Availability with Oracle Database 12c - Technical White Paper

  Read the article

• Designing a software based load balancer

  - by Kishore pandey

  Hello to all Server fault users, I am new to this website but have constantly been using the mother website, stackover flow. Well to begin with, i would like to design a load balancer for the organization i am working for. As i am very new to this whole, idea about load balancing and networks. I am finding it very difficult to start my project. I did a lot of research on already existing load balancer and found some(HAPROXY,NGINX) that could solve my problems, but the point is, I am still in a dilemma if they could answer the following requirements of mine: The client and server in my architecture are distributed. The load balancer should take care of the firewall. LB server should balance the load among all servers present in WWW cloud. The LB server should have some sort of configuration file, with the help of which it is possible to configure the servers. Heart beat: With the help of which it would be possible to check if any server is down, if any server is down the request should be passed to some other server. Various load balancing algorithms of the incoming requests. Easy error handling. It should be fairly possible to prioritize the incoming requests. Is there any already available load balncer solution on the market that could satisfy these requirements? If not is there any base code available with the help of which i could develop my own load balncer. If not where should i start from scratch? I am practically new to everything. Any help from a load balancer expert is very much appreciated. Thanx a ton in advance. Cheers and regards. Kishore

  Read the article

• Which version control should I use for my configuration files?

  - by rakete

  I want to store some of my configuration files (~/.emacs.d/, .Xdefaults, etc. linux $HOME stuff) in version control so I can easily sync them with my notebook/workplace and see my past changes and revert to them should the need arise. So far it seems to me that there are quite some people using git for this and I think that I too want to use a distributed vcs for this (if only to get more used to them) but I can't say that I am very experienced with all things dvcs. I did use darcs and git briefly and so far I can say that I really like the way git handles branches, and I think the possibility to have different branches within the same directory is especially useful for my use case. Darcs on the other hand has cherry picking of patches, which too is quite the convenient feature when managing configuration files (at least I assume it is). So, what would you recommend to use? And what would be your reasoning for your recommendation? What other vcs with nice feature that I haven't mentioned exist and would make a good vcs to store configuration files and why?

  Read the article

• Can I run Excel 2010 on a server?

  - by Glen Little

  This question is not about a person using Excel on a computer that happens to have an Windows Server OS. And it is not about using any Sharepoint services features! The question is about automated processes that use code (Office Automation) to open Excel files, manipulate them, run calculations, read data, save copies of the file and close the files... all in code. In previous versions of Excel the licensing agreement prevented use on a public server, notes from Microsoft warned about the problems trying to use Office Automation in a server environment, and we were warned that Excel was single threaded and not designed for use on a server. Most of the articles about this were written before Office 2010. But now, Excel 2010 is designed to work on a High Performance Computing server using HPC Services for Excel. One HPC document mentions "Windows HPC Server 2008 R2 includes a comprehensive pop-up manager that can handle occasional dialog boxes and pop-up messages". So my question is... is it now "safe" to run code that automates Excel 2010 on a "normal" server without using the HPC services? If not, can the HPC Services for Excel work on a single server? I don't need the high performance, distributed computing, aspect of HPC Services for Excel... just the ability to run Excel on a server. Can that now be done? Thanks, Glen

  Read the article

• Good DBAs Do Baselines

  - by Louis Davidson

  One morning, you wake up and feel funny. You can’t quite put your finger on it, but something isn’t quite right. What now? Unless you happen to be a hypochondriac, you likely drag yourself out of bed, get on with the day and gather more “evidence”. You check your symptoms over the next few days; do you feel the same, better, worse? If better, then great, it was some temporal issue, perhaps caused by an allergic reaction to some suspiciously spicy chicken. If the same or worse then you go to the doctor for some health advice, but armed with some data to share, and having ruled out certain possible causes that are fixed with a bit of rest and perhaps an antacid. Whether you realize it or not, in comparing how you feel one day to the next, you have taken baseline measurements. In much the same way, a DBA uses baselines to gauge the gauge health of their database servers. Of course, while SQL Server is very willing to share data regarding its health and activities, it has almost no idea of the difference between good and bad. Over time, experienced DBAs develop “mental” baselines with which they can gauge the health of their servers almost as easily as their own body. They accumulate knowledge of the daily, natural state of each part of their database system, and so know instinctively when one of their databases “feels funny”. Equally, they know when an “issue” is just a passing tremor. They see their SQL Server with all of its four CPU cores running close 100% and don’t panic anymore. Why? It’s 5PM and every day the same thing occurs when the end-of-day reports, which are very CPU intensive, are running. Equally, they know when they need to respond in earnest when it is the first time they have heard about an issue, even if it has been happening every day. Nevertheless, no DBA can retain mental baselines for every characteristic of their systems, so we need to collect physical baselines too. In my experience, surprisingly few DBAs do this very well. Part of the problem is that SQL Server provides a lot of instrumentation. If you look, you will find an almost overwhelming amount of data regarding user activity on your SQL Server instances, and use and abuse of the available CPU, I/O and memory. It seems like a huge task even to work out which data you need to collect, let alone start collecting it on a regular basis, managing its storage over time, and performing detailed comparative analysis. However, without baselines, though, it is very difficult to pinpoint what ails a server, just by looking at a single snapshot of the data, or to spot retrospectively what caused the problem by examining aggregated data for the server, collected over many months. It isn’t as hard as you think to get started. You’ve probably already established some troubleshooting queries of the type SELECT Value FROM SomeSystemTableOrView. Capturing a set of baseline values for such a query can be as easy as changing it as follows: INSERT into BaseLine.SomeSystemTable (value, captureTime) SELECT Value, SYSDATETIME() FROM SomeSystemTableOrView; Of course, there are monitoring tools that will collect and manage this baseline data for you, automatically, and allow you to perform comparison of metrics over different periods. However, to get yourself started and to prove to yourself (or perhaps the person who writes the checks for tools) the value of baselines, stick something similar to the above query into an agent job, running every hour or so, and you are on your way with no excuses! Then, the next time you investigate a slow server, and see x open transactions, y users logged in, and z rows added per hour in the Orders table, compare to your baselines and see immediately what, if anything, has changed!

  Read the article

• How to organise storage for media content such as video and music?

  - by thor

  Currently, we have a single server hosting all content: music, video and software. This content is downloaded by users through HTTP. Now free space is coming to an end and we are exploring different ways of extending our storage capacity. We want to do it cheap, simple and reliable (protected from disk/ server faults). Currenly, we see two ways: Add a couple of cheap servers with 4 disks (RAID1 ?), run some distributed file-system on top, like GlusterFS. Pros: hopefully, we will see all our disks as single flat file system, just dump content into it and be done. Cons: could be tricky in configuration and handling of faults. Add a couple of cheap servers, all running HTTP servers. Each piece of content (be it a music file or video) is placed on randomly selected two servers. Pros: don't have to deal with RAID, as content is duplicated; single server failure does not bring down any part of content; doubled distribution capacity (as any signle file could be downloaded from any of two servers hosting it). Cons: requires some scripting on part of distribution of content, adding/ removing servers. Do we miss any other ways? Which of the aforementioned options seems to be the best?

  Read the article

• Oracle ties social, CRM, analytics products to customer experience

  - by Richard Lefebvre

  Oracle will embark on a new product strategy that centers on customer experience management, an approach driven by the company’s many recent acquisitions.  The new approach, announced by the company Monday night, will be seen in an expansive suite that features familiar Oracle products -- such as its Fusion CRM platform -- and offerings the company recently gained through acquisitions, including FatWire, RightNow and Vitrue. Billed as Oracle Customer Experience (CX), the suite enables businesses to respond to a market centered on the customer experience, said Anthony Lye, the company’s senior vice president of CRM. Companies “are very aware their products are commoditizing,” Lye said in an interview last week, referring to how the Web and social media channels have empowered customers. Customer experiences start and mature outside of CRM, and applications today need to reflect that shift, Lye said. Businesses thus need to step away from a pure CRM model, he said. Oracle claims CX will improve customer experience management by connecting businesses with customers across Web sites and social channels. Companies can create a single, real-time view of the customer and use predictive analytics of interactions to strengthen the customer experience, Oracle said. “Companies have to connect with their customers wherever, whenever and however they want,” Lye said. “They have to know and understand their customer.” Lye promoted Oracle CX as a suite that will work across channels to complement the company’s applications. A new strategy has been “cooking” for years now, but the acquisitions Oracle has made over the past two years made the time right for a “unique collaboration,” Lye said. CX includes basic Oracle CRM solutions such as Siebel and the new Fusion Apps. It also includes the company’s MDM products, Enterprise Data Quality, Customer Hub and Product Hub. And the suite is rounded out by the services that Oracle recently bought, transactions that created or enhanced the company’s presence in social, marketing, e-commerce and customer service. For instance, FatWire provides tools for marketing. ATG focuses on e-commerce. And RightNow specializes in customer service. Two recent acquisitions -- Collective Intellect and Vitrue -- gave Oracle a seat at the social table. Collective Intellect is a social intelligence program, and Vitrue is a social marketing and engagement platform. Those acquisitions have yet to be finalized. Oracle hopes to eventually integrate the two social offerings, as well as most of the other services, into the CX suite. CX can integrate on Oracle’s standard middleware, and can give users a lower TCO by leveraging it as a single stack on premise or as a cloud solution. Lye deferred questions about the pricing of CX, and instead pitched Oracle’s ability to offer multiple customer experience solutions in one suite. Businesses have struggled with the complexity of infrastructure and modern services that communicate with customers, Lye said. “They’ve struggled to pull all these things together. We’ve done that,” he said. Stephen Powers, a research director at Forrester Research Inc. in Cambridge, Mass., said it’s not surprising for Oracle to offer the CX suite and a related customer experience strategy.  “They’ve got CRM, ATG, FatWire. Clearly, it’s been the strategy for them,” he said. But the challenge for Oracle, and for any other vendor that has gone on an “acquisition spree,” is to connect its many products, Powers said. “The portfolio has to be more than the parts. They’ve got to realize the efficiencies and value of having these pieces to tie them together,” he said. “The proof is in the pudding. Adobe has done a nice job in its space with the products they’ve got. Now, Oracle has got to show it has something.” Albert McKeon (SearchCRM) Published: 25 Jun 2012 : http://searchcrm.techtarget.com/news/2240158644/Oracle-ties-social-CRM-analytics-products-to-customer-experience

  Read the article

• Cisco 3560+ipservices -- IGMP snooping issue with TTL=1

  - by Jander

  I've got a C3560 with Enhanced (IPSERVICES) image, routing multicast between its VLANs with no external multicast router. It's serving a test environment where developers may generate multicast traffic on arbitrary addresses. Everything is working fine except when someone sends out multicast traffic with TTL=1, in which case the multicast packet suppression fails and the traffic is broadcast to all members of the VLAN. It looks to me like because the TTL is 1, the multicast routing subsystem doesn't see the packets, so it doesn't create a mroute table entry. If I send out packets with TTL=2 briefly, then switch to TTL=1 packets, they are filtered correctly until the mroute entry expires. My question: is there some trick to getting the switch to filter the TTL=1 packets, or am I out of luck? Below are the relevant parts of the config, with a representative VLAN interface. I can provide more info as needed. #show run ... ip routing ip multicast-routing distributed no ip igmp snooping report-suppression ! interface Vlan44 ip address 172.23.44.1 255.255.255.0 no ip proxy-arp ip pim passive ... #show ip igmp snooping vlan 44 Global IGMP Snooping configuration: ------------------------------------------- IGMP snooping : Enabled IGMPv3 snooping (minimal) : Enabled Report suppression : Disabled TCN solicit query : Disabled TCN flood query count : 2 Robustness variable : 2 Last member query count : 2 Last member query interval : 1000 Vlan 44: -------- IGMP snooping : Enabled IGMPv2 immediate leave : Disabled Multicast router learning mode : pim-dvmrp CGMP interoperability mode : IGMP_ONLY Robustness variable : 2 Last member query count : 2 Last member query interval : 1000

  Read the article

• EventID 1058 Code 5, Sysvol is subdir of Sysvol - how to fix?

  - by nulliusinverba

  I have been trying to resolve this error, like many others: The processing of Group Policy failed. Windows attempted to read the file \domain.local\SysVol\domain.local\Policies{3EF90CE1-6908-44EC-A750-F0BA70548600}\gpt.ini from a domain controller and was not successful. Group Policy settings may not be applied until this event is resolved. This issue may be transient and could be caused by one or more of the following: a) Name Resolution/Network Connectivity to the current domain controller. b) File Replication Service Latency (a file created on another domain controller has not replicated to the current domain controller). c) The Distributed File System (DFS) client has been disabled. Error code: 5 = Access Denied. The incredibly helpful post is this one (http://www.experts-exchange.com/OS/Microsoft_Operating_Systems/Server/2003_Server/A_1073-Diagnosing-and-repairing-Events-1030-and-1058.html). Quoting from this post: HERE IS A LIST OF POTENTIAL PROBLEMS THAT CAN LEAD TO 1030 AND 1058 EVENT ERRORS: --Sometimes the permissions of the file folders that contain Group policies (the Sysvol folder) can be corrupted. --Sometimes you have problems with NetBIOS: --Sometimes the GPO itself is corrupt, or you have a partial set of data for that GPO. --Sometimes you may have problems with File Replication Services, which almost always indicates a problem with DNS --Sysvol may be a subfolder of itself: Sysvol/Sysvol I have the problem listed where sysvol is a subfolder of sysvol. The directory structure is: -sysvol -domain -staging -staging areas -sysvol (shared as "\\server\sysvol") -domain.local -ClientAgent -Policies -scripts Interestingly, the second sysvol folder is the one that is shared as "\server\sysvol". This makes me confident this is the issue with the permissions and error code 5. Also interestingly, my server 2008 R2 servers can see it fine - my server 2008 servers cannot, and get the error. This is consistent across all my servers. This latter fact makes me uncertain what I need to do to fix this up. Do I, e.g., simply move the shared sysvol folder up a level to replace the non-shared one? Any help greatly appreciated. Cheers, Tim.

  Read the article

• ARTS Reference Model for Retail

  - by Sanjeev Sharma

  Consider a hypothetical scenario where you have been tasked to set up retail operations for a electronic goods or daily consumables or a luxury brand etc. It is very likely you will be faced with the following questions: What are the essential business capabilities that you must have in place?  What are the essential business activities under-pinning each of the business capabilities, identified in Step 1? What are the set of steps that you need to perform to execute each of the business activities, identified in Step 2? Answers to the above will drive your investments in software and hardware to enable the core retail operations. More importantly, the choices you make in responding to the above questions will several implications in the short-run and in the long-run. In the short-term, you will incur the time and cost of defining your technology requirements, procuring the software/hardware components and getting them up and running. In the long-term, as you grow in operations organically or through M&A, partnerships and franchiser business models  you will invariably need to make more technology investments to manage the greater complexity (scale and scope) of business operations.  "As new software applications, such as time & attendance, labor scheduling, and POS transactions, just to mention a few, are introduced into the store environment, it takes a disproportionate amount of time and effort to integrate them with existing store applications. These integration projects can add up to 50 percent to the time needed to implement a new software application and contribute significantly to the cost of the overall project, particularly if a systems integrator is called in. This has been the reality that all retailers have had to live with over the last two decades. The effect of the environment has not only been to increase costs, but also to limit retailers' ability to implement change and the speed with which they can do so." (excerpt taken from here) Now, one would think a lot of retailers would have already gone through the pain of finding answers to these questions, so why re-invent the wheel? Precisely so, a major effort began almost 17 years ago in the retail industry to make it less expensive and less difficult to deploy new technology in stores and at the retail enterprise level. This effort is called the Association for Retail Technology Standards (ARTS). Without standards such as those defined by ARTS, you would very likely end up experiencing the following: Increased Time and Cost due to resource wastage arising from re-inventing the wheel i.e. re-creating vanilla processes from scratch, and incurring, otherwise avoidable, mistakes and errors by ignoring experience of others Sub-optimal Process Efficiency due to narrow, isolated view of processes thereby ignoring process inter-dependencies i.e. optimizing parts but not the whole, and resulting in lack of transparency and inter-departmental finger-pointing Embracing ARTS standards as a blue-print for establishing or managing or streamlining your retail operations can benefit you in the following ways: Improved Time-to-Market from parity with industry best-practice processes e.g. ARTS, thus avoiding “reinventing the wheel” for common retail processes and focusing more on customizing processes for differentiations, and lowering integration complexity and risk with a standardized vocabulary for exchange between internal and external i.e. partner systems Lower Operating Costs by embracing the ARTS enterprise-wide process reference model for developing and streamlining retail operations holistically instead of a narrow, silo-ed view, and  procuring IT systems in compliance with ARTS thus avoiding IT budget marginalization While parity with industry standards such as ARTS business process model by itself does not create a differentiation, it does however provide a higher starting point for bridging the strategy-execution gap in setting up and improving retail operations.

  Read the article

• Multiple servers vs 1 big server performace

  - by pistacchio

  Hi to all! My team of developers has suggested a server structure for an upcoming project we are developing. Our structure is "logical", meaning that the various logical components of the application (it is a distributed one) relies on different servers. Some components are more critical than others and will be subjected to more load. Our proposal was to have 1 server per component but the hardware guys suggested to replace the various machines with a single, bigger one with virtual servers. They're gonna use Blade Servers. Now, I'm not an expert at all, but my question to the guys was: so if we need, for example, 3 2GHz CPU / 2GB RAM machines and you give me 1 machine with 3 2GHz CPUs and 6 GB of RAM it is the same? They told me it is. Is this accurate? What are the advantages or disadvantages of both the solutions? What are the generally accepted best practices? Could you point out some URL reference dealing with the problem? Thank you in advance! EDIT: Some more info. The (internet / intranet) application is already layered. We have some servers on the DMZ that will expose pages to the internet and the databases are on their own machines. What we want to split (and they want to join) are some webservers that mainly expose webservices. One is a DAL that communicates with the database layer, one is our Single Sign On / User Profile application that gets called once per page and one is a clone of what seen on the Internet to be used on our lan.

  Read the article

• Proving What You are Worth

  - by Ted Henson

  Here is a challenge for everyone. Just about everyone has been asked to provide or calculate the Return on Investment (ROI), so I will assume everyone has a method they use. The problem with stopping once you have an ROI is that those in the C-Suite probably do not care about the ROI as much as Return on Equity (ROE). Shareholders are mostly concerned with their return on the money the invested. Warren Buffett looks at ROE when deciding whether to make a deal or not. This article will outline how you can add more meaning to your ROI and show how you can potentially enhance the ROE of the company.   First I want to start with a base definition I am using for ROI and ROE. Return on investment (ROI) and return on equity (ROE) are ways to measure management effectiveness, parts of a system of measures that also includes profit margins for profitability, price-to-earnings ratio for valuation, and various debt-to-equity ratios for financial strength. Without a set of evaluation metrics, a company's financial performance cannot be fully examined by investors. ROI and ROE calculate the rate of return on a specific investment and the equity capital respectively, assessing how efficient financial resources have been used. Typically, the best way to improve financial efficiency is to reduce production cost, so that will be the focus. Now that the challenge has been made and items have been defined, let’s go deeper. Most research about implementation stops short at system start-up and seldom addresses post-implementation issues. However, we know implementation is a continuous improvement effort, and continued efforts after system start-up will influence the ultimate success of a system.   Most UPK ROI’s I have seen only include the cost savings in developing the training material. Some will also include savings based on reduced Help Desk calls. Using just those values you get a good ROI. To get an ROE you need to go a little deeper. Typically, the best way to improve financial efficiency is to reduce production cost, which is the purpose of implementing/upgrading an enterprise application. Let’s assume the new system is up and running and all users have been properly trained and are comfortable using the system. You provide senior management with your ROI that justifies the original cost. What you want to do now is develop a good base value to a measure the current efficiency. Using usage tracking you can look for various patterns. For example, you may find that users that are accessing UPK assistance are processing a procedure, such as entering an order, 5 minutes faster than those that don’t.  You do some research and discover each minute saved in processing a claim saves the company one dollar. That translates to the company saving five dollars on every transaction. Assuming 100,000 transactions are performed a year, and all users improve their performance, the company will be saving $500,000 a year. That $500,000 can be re-invested, used to reduce debt or paid to the shareholders.   With continued refinement during the life cycle, you should be able to find ways to reduce cost. These are the type of numbers and productivity gains that senior management and shareholders want to see. Being able to quantify savings and increase productivity may also help when seeking a raise or promotion.

  Read the article

• Synchronize two directories on linux pc

  - by Gab

  I need a distributed filesystem (or a synchronization tool) that is capable of keeping a directory synchronized across 4 pc. My requirements are: offline access (data must be available offline on each pc) preserve execution rights: some files are marked executable on a linux partition. This flag should be replicated. efficient sync strategy: some of my files are 20GB, they are changed quite often, but in very little parts (Virtualbox images). Delta transmissions are welcome. efficient handling of space: no history for files, files shouldn't be copied to temp directories "just in case you break it". it must propagate deletions of files modification can happen in any of the 4 pcs, they should be propagated when other pc are connected. Other specs of my solution are: Sync is over a lan, the total amount of data to be synced is around 180GB, in some ten thousand files. Changes are small, but can happen in big files. At the moment i'm interested in a linux only solution. conflicts either don't happen or are solved with "last one wins" I haven't found any good solution. I've been trying: unison: it is the only one working at the moment, but during the hashing phase it hangs my pc for some minute, disk light steady on. Sparkleshare doesn't handle large files nicely. It keeps an history of all your changes that grows up indefinitely. They promise it will be fixed in next releases, but at the moment it still doesn't fit my needs. owncloud (keeps history of each file i change) coda ? (help! i couldn't set it up correctly!) git-annex assistant transforms all your files in symlinks and mark the original file as read only ("just in case you make a mistake while you modify it"!). Before you edit a file you have to issue a special command "git-annex unlock", that creates a local copy of the file, and you have to remember to lock it again if you want it synchronized. What to try next?

  Read the article

• Extract sender activity from postfix logs for auditing user

  - by Aseques

  We have a mail user on our postfix server that was using the company mail to send compromising information to the competence. I've been asked to make a report of the actions for that user in the last time. There are tools like pflogsumm and others that can extract statistic data, but I haven't so far find anything useful to get all the info for a user because the data is in multiple lines. I'd like to get something like this: For the sent mail 11/11/11 00:00:00 [email protected] -> [email protected] 11/11/11 00:00:01 [email protected] -> [email protected] For the received mail 10/10/11 00:00:00 [email protected] -> [email protected] 10/10/11 00:00:01 [email protected] -> [email protected] I know I can do a script by myself, but matching the postfix ID for every mail is not something that can be made with a simple grep, and I've a lot of mail history that I have to recheck distributed among diferent files and so on. The source log is the standard postfix format, for example this one... Sep 13 16:15:57 server postfix/qmgr[18142]: B35CB5ED3D: from=<[email protected], size=10755, nrcpt=1 (queue active) Sep 13 16:15:57 server postfix/smtpd[32099]: disconnect from localhost[127.0.0.1] Sep 13 16:15:57 server postfix/smtp[32420]: 58C3E5EC9C: to=<[email protected]>, relay=127.0.0.1[127.0.0.1]:10024, delay=1.4, delays=0.01/0/0/1.4, dsn=2.0.0, status=sent (250 2.0.0 Ok, id=32697-04, from MTA([127.0.0.1]:10025): 250 2.0.0 Ok: queued as B35CB5ED3D) Sep 13 16:15:57 server postfix/qmgr[18142]: 58C3E5EC9C: removed Sep 13 16:15:57 server postfix/smtp[32379]: B35CB5ED3D: to=<[email protected]>, relay=mail.anothercompany.com[123.123.123.163]:25, delay=0.06, delays=0.03/0/0.01/0.02, dsn=2.0.0, status=sent (250 2.0.0 Ok: queued as 77D0EB6C025) Sep 13 16:15:57 server postfix/qmgr[18142]: B35CB5ED3D: removed

  Read the article

• Large, high performance object or key/value store for HTTP serving on Linux

  - by Tommy

  I have a service that serves images to end users at a very high rate using plain HTTP. The images vary between 4 and 64kbytes, and there are 1.300.000.000 of them in total. The dataset is about 30TiB in size and changes (new objects, updates, deletes) make out less than 1% of the requests. The number of requests pr. second vary from 240 to 9000 and is dispersed pretty much all over, with few objects being especially "hot". As of now, these images are files on a ext3 filesystem distributed read only across a large amount of mid range servers. This poses several problems: Using a fileysystem is very inefficient since the metadata size is large, the inode/dentry cache is volatile on linux and some daemons tend to stat()/readdir() it's way through the directory structure, which in my case becomes very expensive. Updating the dataset is very time consuming and requires remounting between set A and B. The only reasonable handling is operating on the block device for backup, copying, etc. What I would like is a deamon that: speaks HTTP (get, put, delete and perhaps update) stores data it in an efficient structure. The index should remain in memory, and considering the amount of objects, the overhead must be small. The software should be able to handle massive connections with slow (if any) time needed to ramp up. Index should be read in memory at startup. Statistics would be nice, but not mandatory. I have experimented a bit with riak, redis, mongodb, kyoto and varnish with persistent storage, but I haven't had the chance to dig in really deep yet.

  Read the article

< Previous Page | 89 90 91 92 93 94 95 96 97 98 99 100  | Next Page >