Search Results

Search found 5011 results on 201 pages for 'grand master t'.

Page 40/201 | < Previous Page | 36 37 38 39 40 41 42 43 44 45 46 47  | Next Page >

  • Rebasing a branch which is public

    - by Dror
    I'm failing to understand how to use git-rebase, and I consider the following example. Let's start a repository in ~/tmp/repo: $ git init Then add a file foo $ echo "hello world" > foo which is then added and committed: $ git add foo $ git commit -m "Added foo" Next, I started a remote repository. In ~/tmp/bare.git I ran $ git init --bare In order to link repo to bare.git I ran $ git remote add origin ../bare.git/ $ git push --set-upstream origin master Next, lets branch, add a file and set an upstream for the new branch b1: $ git checkout -b b1 $ echo "bar" > foo2 $ git add foo2 $ git commit -m "add foo2 in b1" $ git push --set-upstream origin b1 Now it is time to switch back to master and change something there: $ echo "change foo" > foo $ git commit -a -m "changed foo in master" $ git push At this point in master the file foo contain changed foo, while in b1 it is still hello world. Finally, I want to sync b1 with the progress made in master. $ git checkout b1 $ git fetch origin $ git rebase origin/master At this point git st returns: # On branch b1 # Your branch and 'origin/b1' have diverged, # and have 2 and 1 different commit each, respectively. # (use "git pull" to merge the remote branch into yours) # nothing to commit, working directory clean At this point the content of foo in the branch b1 is change foo as well. So what does this warning mean? I expected I should do a git push, git suggests to do git pull... According to this answer, this is more or less it, and in his comment @FrerichRaabe explicitly say that I don't need to do a pull. What's going on here? What is the danger, how should one proceed? How should the history be kept consistent? What is the interplay between the case described above and the following citation: Do not rebase commits that you have pushed to a public repository. taken from pro git book. I guess it is somehow related, and if not I would love to know why. What's the relation between the above scenario and the procedure I described in this post.

    Read the article

  • How to take mysql replication backup

    - by user53864
    I have a MySQL master-master replication setup with a slave for each master(only one master used for read/writes at a time) on Ubuntu server. Wondering what would be the best way to schedule backup of replication databases with mysqldump. I have following clarifications because of which could not proceed further. Scheduling mysqldump backup on masters safe for replication? Connecting masters with GUI applications(workbench) for database manipulations(read, writes.. by developers) is safe? Any inputs are welcome.

    Read the article

  • asp.net Can I force every page to inherit from a base page? Also should some of this logic be in my master page?

    - by Bex
    Hi! I have a web app that has a base page. Each page needs to inherit from this base page as it contains properties they all need as well as dealing with the login rights. My base page has some properties, eg: IsRole1, IsRole2, currentUserID, Role1Allowed, Role2Allowed. On the init of each page I set the properties "Role1Allowed" and "Role2Allowed" Private Sub Page_Init(ByVal sender As Object, ByVal e As System.EventArgs) Handles Me.Init Role1Allowed = True Role2Allowed= False End Sub The basepage then decides if the user needs redirecting. 'Sample code so not exactly what is going to be, bug gives the idea Protected Overridable Sub Page_Load(ByVal sender As Object, ByVal e As System.EventArgs) If Role1Allowed And Not Role1 Then 'Redirect somewhere End If End Sub The page then must override this pageload if they need anything else in it, but making sure they call the base pageload first. Protected Overrides Sub Page_Load(ByVal sender As Object, ByVal e As System.EventArgs) Handles Me.Load MyBase.Page_Load(sender, e) If Not IsPostBack Then BindGrid() End If End Sub The other properties (IsRole1, IsRole, currentUserID) are also accessible by the page so it can be decided if certain things need doing based on the user. (I hope this makes sense) Ok so I have 2 questions Should this functionality be in the base page or should it somehow be in the master, and if so how would I get access to all the properties if it was? As there are multiple people working on this project and creating pages some are forgetting to inherit from this basepage, or call the base pageload when overriding it. Is there any way to force them to do this? Thanks for any help. bex

    Read the article

  • Start Daemonised GNU Screen from script a allow calling script to end

    - by tez
    I have a script on an embedded device that calls screen to start if a user logs in via a ssh session... #!/bin/sh SCREENRUNNING=`pgrep SCREEN` if [ -z "$SCREENRUNNING" ]; then echo "Screen not running so let's start the Master session sleep 2 screen -dmS Master sleep 2 screen -x root/Master else echo "Screen is already running let's connect to existing session" sleep 2 screen -x root/Master fi However this keeps the calling script active till the screen session exits,even if it's detached. What I want to do is have the calling script finish and exit while the screen session stays active. I've tried daemonising the screen -x lines and adding an & to the end of the screen -x lines neither of which work properly. Ideas?

    Read the article

  • BIND: forward 1st level zone

    - by raven
    First of all: sorry for the language, English is not my primary language. I have star-like DNS structure with many filials (more that 2): ^ | v filialNS_1.filial_1.city.local <---- ns.main.city.local <---- filialNS_2.filial_2.city.local ^ | v ns.mail.city.local is slave of all filials zones filialNS_1 is master of filial_1.city.local filialNS_2 is master of filial_2.city.local filialNS_N is master of filial_N.city.local I want to: serve DNS queries for xxx.filial_N.city.local with filialNS_N.filial_N.city.local forward all queries for xxx.xxx.xxx.local from filialNS_N to ns.main.city.local forward other queries to our provider's DNS on filial (or google-public-dns or anything else) FILIAL CONFIG named.conf zone "filial_1.city.local" { type master; file "/etc/namedb/dynamic/filial_1.city.local"; allow-update { key DHCP_UPDATER; }; allow-transfer { <ns.main.city.local IP address> }; }; zone "2.76.10.in-addr.arpa" { type master; file "/etc/namedb/dynamic/2.76.10.in-addr.arpa"; allow-update { key DHCP_UPDATER; }; allow-transfer { <ns.main.city.local IP address> }; }; zone "local." { type forward; forward only; forwarders { <ns.main.city.local IP address> }; }; nslookup server.filial_1.city.local - works fine nslookup server.main.city.local Server: 127.0.0.1 Address: 127.0.0.1#53 ** server can't find server.main.city.local: NXDOMAIN Where am I going wrong?

    Read the article

  • Can you help me understand my SATA/RAID options?

    - by andrz_001
    I've a gigabyte GA-M720-US3 motherboard. Recently, I noticed the following during boot: IDE channel 0 Master (none) IDE channel 0 Slave (none) IDE channel 2 Master (my hdd) IDE channel 2 Slave (my dvd drive) IDE channel 3 Master (none) IDE channel 3 Slave (none) Of course, the same information is contained in the BIOS/CMOS. The HDD is connected to the mobo via a SATA(2?) cable at the port(?) labeled SATA2_0. The DVD drive is connected by a similar cable at SATA2_1. Why doesn't the information displayed during the boot and in BIOS reflect how I plugged the cables in? I mean, why "none" for channel 0 when there is something in SATA2_0. (or is that serious naivete on my part!?) Where's Channel 1 master and slave? Since these are SATA cables and not the IDE ribbons from a time ago, why the whole master/slave declaration during boot and in BIOS? Should my BIOS reflect the fact that these are SATA cables? I mean, in BIOS, should the "Onchip SATA mode IDE" be set to RAID or AHCI instead of IDE? Any replies, answers, suggestions, links, tips will be much appreciated. Thank you in advance!

    Read the article

  • What is the fastest method to restore MySQL replication?

    - by dwhere
    I have a MySQL (5.1) master-slave replication pair and replication to the slave has failed. It failed because the master ran out of disk space and the relay-logs became corrupt. The master is now back online and working properly. Since there is this error in the log the slave process can't simply be restarted. The server has a single 40GB InnoDB database and I would like to know what is the fastest method for getting the slave back in sync to minimize downtime.

    Read the article

  • Múltiples clientes en el mismo Oracle UCM

    - by [email protected]
    Estamos muy activos con la implantación de plataformas ECM que den servicio a múltiples clientes. Consiguiendo 2 objetivos muy importantes:El cliente final puede pagar al proveedor por una plataformas ECM como servicio (SaaS). Y, lógicamente, se ahorra en complejidad y gastos de infraestructura, administración, formación, almacenamiento, etc...Hemos estado explicando estos días el modelo Master-Proxy de Oracle UCM con el que podemos implantar este tipo de plataformas. No siempre será la solución más adecuada porque a veces vamos a querer disponer de plataformas compartidas, pero con clientes completamente aislados. Siempre, la consola de migración nos permite exportar e importar componentes, metadatos, contenidos, workflows, etc... para que elijamos el modelo más adecuado para cada caso.Pero, ¿Cómo funciona?. Podéis ver en la imágen que se basa en la instalación de varias instancias de UCM y configurarlas de forma que varias de ellas se comporten como "Master" (digo varias para conseguir alta disponibilidad), y el resto se comporten como "Proxy" (también varias instancias de UCM pueden comportarse como un mismo "Proxy" permitiendo balancear la carga en función de que cada cliente requiera más o menos rendimiento). Esta configuración (que vemos en la imágen adjunta), nos permite:Delegar la gestión de usuarios de cada cliente. Los usuarios del Master podrán acceder a todos los Proxies, pero los usuarios de cada proxy sólo acceden a su repositorio.Delegar funcionalidad y componentes. Es posible configurar diferentes funcionalidades en cada proxy de forma que algunos servidores estén especializados en Web Content Management, otros en Document Management (por ejemplo).Diferentes modelos de metadatos. Podemos modelar unos tipos documentales generales para toda la plataforma y otros particulares diferentes en cada UCM "Proxy".Conseguir una centralización de búsquedas y acceso a repositorios de documentación con diferentes juegos de caracteres. Un UCM Master puede centralizar la búsqueda en UCM's proxy que alberguen documentación en diferentes juegos de caracteres (por ejemplo un UCM para documentación de idiomas "Western European" (inglés, español, francés, alemán,...) y otro UCM proxy bajo juego de caracteres "Asian" (japones, coreano, chino,...).Fuentes:Toda la información detallada se encuentra en la documentación de Oracle UCM, aquí:http://download.oracle.com/docs/cd/E10316_01/ouc.htmY en concreto, lo relativo a plataformas, en el documento "Planning and Implementation Guide", aquí:plan_implement_guide_10en.pdf

    Read the article

  • Reading and conditionally updating N rows, where N > 100,000 for DNA Sequence processing

    - by makerofthings7
    I have a proof of concept application that uses Azure tables to associate DNA sequences to "something". Table 1 is the master table. It uniquely lists every DNA sequence. The PK is a load balanced hash of the RK. The RK is the unique encoded value of the DNA sequence. Additional tables are created per subject. Each subject has a list of N DNA sequences that have one reference in the Master table, where N is 100,000. It is possible for many tables to reference the same DNA sequence, but in this case only one entry will be present in the Master table. My Azure dilemma: I need to lock the reference in the Master table as I work with the data. I need to handle timeouts, and prevent other threads from overwriting my data as one C# thread is working with the information. Other threads need to realise that this is locked, and move onto other unlocked records and do the work. Ideally I'd like to get some progress report of how my computation is going, and have the option to cancel the process (and unwind the locks). Question What is the best approach for this? I'm looking at these code snippets for inspiration: http://blogs.msdn.com/b/jimoneil/archive/2010/10/05/azure-home-part-7-asynchronous-table-storage-pagination.aspx http://stackoverflow.com/q/4535740/328397

    Read the article

  • Backup those keys, citizen

    - by BuckWoody
    Periodically I back up the keys within my servers and databases, and when I do, I blog a reminder here. This should be part of your standard backup rotation – the keys should be backed up often enough to have at hand and again when they change. The first key you need to back up is the Service Master Key, which each Instance already has built-in. You do that with the BACKUP SERVICE MASTER KEY command, which you can read more about here. The second set of keys are the Database Master Keys, stored per database, if you’ve created one. You can back those up with the BACKUP MASTER KEY command, which you can read more about here. Finally, you can use the keys to create certificates and other keys – those should also be backed up. Read more about those here. Anyway, the important part here is the backup. Make sure you keep those keys safe! Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • MaaS minimum requirements with juju-jitsu?

    - by Christopher Shen Mu Long
    I've browsed through so many different sites and found so much contradictory information. As I am getting tired of this and do belive this question affects many other users, so I would like to collect the "once and for all times" answer. Unfortunately, the documentation on MaaS and Juju is ... well, not the best, sorry to say that. What are the minimum system requirements for setting up a MaaS cluster, which is going to be orchestrated with juju-jitsu? Do they need to have the exact system specifications or can I just combine different hardware? What are the minimum requirements for the master machine? E.g. "You need at least 8GB of RAM, a dual core CPU with at least 3.0 GHz." How many machines to I need to deploy MaaS on? I've read six machines, nine machines, and so on. I clearly want to know: "You need one for the Master and e.g. five nodes." Do I need to attach as many NICs (network interface cards) to my master machine as there are nodes, or can I simply attach two NICs and a switch? One NIC for connecting to the internet, one for handling the MaaS tasks, connected to a switch, which connects my nodes to the master? Is Juju now ready for local deployment? The last time I experimented with juju and had to reboot my machine, the services orchestrated by juju were gone. This was an issue I also found on the official juju site. Unfortunately, as mentioned above, the documentation is not the best, so I could not find the necessary info on that again. So: Can I use juju on a local environment or will a reboot break my setup?

    Read the article

  • Loadbalancing Questions

    - by Van Holtz
    I have been learning networking for about 4 months. Wrote a single standalone Multiplayer server and succeeded with authoritative approach. Now I want to extend it by splitting the single server into clusters to allow even more players to log in to avoid latency issues. Now I have protyped the Loadbalancing server and its running pretty good so far. This is my architecture, I have a master server which acts as a proxy, every sub servers(chat, login, game) connect to the master server as well as all the clients. when a client connects, Client Request: Send Request - MS(Master) - Decides which SS(SubServer) to forward to - Forwards Request to SS - SS - Analyze Message - Send Response to MS - Decides which Client to forward to - Forwards Response to Client Well, it looks like its going through lots of stages. it takes double the time to process the message than a single server approach. i feel like my model isnt the best or i may be wrong. is there any better model or the one they use in professional games? I still want a Master-SubServer approach. I just want to clarify that I'm going in the right direction before writing all my codes. Thanks for any answer :)

    Read the article

  • Maintaining Two Separate Software Versions From the Same Codebase in Version Control

    - by Joseph
    Let's say that I am writing two different versions of the same software/program/app/script and storing them under version control. The first version is a free "Basic" version, while the second is a paid "Premium" version that takes the codebase of the free version and expands upon it with a few extra value-added features. Any new patches, fixes, or features need to find their way into both versions. I am currently considering using master and develop branches for the main codebase (free version) along side master-premium and develop-premium branches for the paid version. When a change is made to the free version and merged to the master branch (after thorough testing on develop of course), it gets copied over to the develop-premium branch via the cherry-pick command for more testing and then merged into master-premium. Is this the best workflow to handle this situation? Are there any potential problems, caveats, or pitfalls to be aware of? Is there a better branching strategy than what I have already come up with? Your feedback is highly appreciated! P.S. This is for a PHP script stored in Git, but the answers should apply to any language or VCS.

    Read the article

  • Complex string matching with fuzzywuzzy

    - by That1Guy
    I'm attempting to write a process that matches obscure strings to a single 'master string' for further processing. I have a lot of data that looks something like this: Basketball Basket Ball Football BasketBallR BBall BBall - r FootB ...and so on. These need to be mapped to a master record like so: Basketball = Basket Ball, BBall Basketball - R = BasketBallR, BBall - r I also have instances of data resembling this format: Football -r FootBall - r-g/H,Q,HH These situations need to be separated into different categories before being mapped. For example FootBall - r-g/H,Q,HH should be: Football - r Football - g Football - H Football - Q Football - HH At this point, it still needs to be mapped to a master record... I've tried several different combinations of fuzzywuzzy matching methods, Levenshtein Distance measurements, regex, etc. and can't seem to find a reliable method to logically associate different naming styles of a single item with a master name. I'm throwing my hands up in desperation. Are there any existing python resources than can help sort out my problem? Are there other options? Can anybody point out an obvious option that I might have overlooked? Basically, any suggestion, solution, resource or alternative method is greatly appreciated.

    Read the article

  • Does Ivy's url resolver support transitive retrieval?

    - by Sean
    For some reason I can't seem to resolve the dependencies of my dependencies when using a url resolver to specify a repository's location. However, when using the ibiblio resolver, I am able to retrieve them. For example: <!-- Ivy File --> <ivy-module version="1.0"> <info organisation="org.apache" module="chained-resolvers"/> <dependencies> <dependency org="commons-lang" name="commons-lang" rev="2.0" conf="default"/> <dependency org="checkstyle" name="checkstyle" rev="5.0"/> </dependencies> </ivy-module> <!-- ivysettings file --> <ivysettings> <settings defaultResolver="chained"/> <resolvers> <chain name="chained"> <url name="custom-repo"> <ivy pattern="http://my.internal.domain.name/ivy/[organisation]/[module]/[revision]/ivy-[revision].xml"/> <artifact pattern="http://my.internal.domain.name/ivy/[organisation]/[module]/[revision]/[artifact]-[revision].[ext]"/> </url> <url name="ibiblio-mirror" m2compatible="true"> <artifact pattern="http://mirrors.ibiblio.org/pub/mirrors/maven2/[organisation]/[module]/[revision]/[artifact]-[revision].[ext]" /> </url> <ibiblio name="ibiblio" m2compatible="true"/> </chain> </resolvers> </ivysettings> <!-- checkstyle ivy.xml file generated from pom via ivy:install task --> <?xml version="1.0" encoding="UTF-8"?> <ivy-module version="1.0" xmlns:m="http://ant.apache.org/ivy/maven"> <info organisation="checkstyle" module="checkstyle" revision="5.0" status="release" publication="20090509202448" namespace="maven2" > <license name="GNU Lesser General Public License" url="http://www.gnu.org/licenses/lgpl.txt" /> <description homepage="http://checkstyle.sourceforge.net/"> Checkstyle is a development tool to help programmers write Java code that adheres to a coding standard </description> </info> <configurations> <conf name="default" visibility="public" description="runtime dependencies and master artifact can be used with this conf" extends="runtime,master"/> <conf name="master" visibility="public" description="contains only the artifact published by this module itself, with no transitive dependencies"/> <conf name="compile" visibility="public" description="this is the default scope, used if none is specified. Compile dependencies are available in all classpaths."/> <conf name="provided" visibility="public" description="this is much like compile, but indicates you expect the JDK or a container to provide it. It is only available on the compilation classpath, and is not transitive."/> <conf name="runtime" visibility="public" description="this scope indicates that the dependency is not required for compilation, but is for execution. It is in the runtime and test classpaths, but not the compile classpath." extends="compile"/> <conf name="test" visibility="private" description="this scope indicates that the dependency is not required for normal use of the application, and is only available for the test compilation and execution phases." extends="runtime"/> <conf name="system" visibility="public" description="this scope is similar to provided except that you have to provide the JAR which contains it explicitly. The artifact is always available and is not looked up in a repository."/> <conf name="sources" visibility="public" description="this configuration contains the source artifact of this module, if any."/> <conf name="javadoc" visibility="public" description="this configuration contains the javadoc artifact of this module, if any."/> <conf name="optional" visibility="public" description="contains all optional dependencies"/> </configurations> <publications> <artifact name="checkstyle" type="jar" ext="jar" conf="master"/> </publications> <dependencies> <dependency org="antlr" name="antlr" rev="2.7.6" force="true" conf="compile->compile(*),master(*);runtime->runtime(*)"/> <dependency org="apache" name="commons-beanutils-core" rev="1.7.0" force="true" conf="compile->compile(*),master(*);runtime->runtime(*)"/> <dependency org="apache" name="commons-cli" rev="1.0" force="true" conf="compile->compile(*),master(*);runtime->runtime(*)"/> <dependency org="apache" name="commons-logging" rev="1.0.3" force="true" conf="compile->compile(*),master(*);runtime->runtime(*)"/> <dependency org="com.google.collections" name="google-collections" rev="0.9" force="true" conf="compile->compile(*),master(*);runtime->runtime(*)"/> </dependencies> </ivy-module> Using the "ibiblio" resolver I have no problem resolving my project's two dependencies (commons-lang 2.0 and checkstyle 5.0) and checkstyle's dependencies. However, when attempting to exclusively use the "custom-repo" or "ibiblio-mirror" resolvers, I am able to resolve my project's two explicitly defined dependencies, but not checkstyle's dependencies. Is this possible? Any help would be greatly appreciated.

    Read the article

  • Benchmarking MySQL Replication with Multi-Threaded Slaves

    - by Mat Keep
    0 0 1 1145 6530 Homework 54 15 7660 14.0 Normal 0 false false false EN-US JA X-NONE /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-parent:""; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:Cambria; mso-ascii-font-family:Cambria; mso-ascii-theme-font:minor-latin; mso-hansi-font-family:Cambria; mso-hansi-theme-font:minor-latin; mso-ansi-language:EN-US;} The objective of this benchmark is to measure the performance improvement achieved when enabling the Multi-Threaded Slave enhancement delivered as a part MySQL 5.6. As the results demonstrate, Multi-Threaded Slaves delivers 5x higher replication performance based on a configuration with 10 databases/schemas. For real-world deployments, higher replication performance directly translates to: · Improved consistency of reads from slaves (i.e. reduced risk of reading "stale" data) · Reduced risk of data loss should the master fail before replicating all events in its binary log (binlog) The multi-threaded slave splits processing between worker threads based on schema, allowing updates to be applied in parallel, rather than sequentially. This delivers benefits to those workloads that isolate application data using databases - e.g. multi-tenant systems deployed in cloud environments. Multi-Threaded Slaves are just one of many enhancements to replication previewed as part of the MySQL 5.6 Development Release, which include: · Global Transaction Identifiers coupled with MySQL utilities for automatic failover / switchover and slave promotion · Crash Safe Slaves and Binlog · Optimized Row Based Replication · Replication Event Checksums · Time Delayed Replication These and many more are discussed in the “MySQL 5.6 Replication: Enabling the Next Generation of Web & Cloud Services” Developer Zone article  Back to the benchmark - details are as follows. Environment The test environment consisted of two Linux servers: · one running the replication master · one running the replication slave. Only the slave was involved in the actual measurements, and was based on the following configuration: - Hardware: Oracle Sun Fire X4170 M2 Server - CPU: 2 sockets, 6 cores with hyper-threading, 2930 MHz. - OS: 64-bit Oracle Enterprise Linux 6.1 - Memory: 48 GB Test Procedure Initial Setup: Two MySQL servers were started on two different hosts, configured as replication master and slave. 10 sysbench schemas were created, each with a single table: CREATE TABLE `sbtest` (    `id` int(10) unsigned NOT NULL AUTO_INCREMENT,    `k` int(10) unsigned NOT NULL DEFAULT '0',    `c` char(120) NOT NULL DEFAULT '',    `pad` char(60) NOT NULL DEFAULT '',    PRIMARY KEY (`id`),    KEY `k` (`k`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1 10,000 rows were inserted in each of the 10 tables, for a total of 100,000 rows. When the inserts had replicated to the slave, the slave threads were stopped. The slave data directory was copied to a backup location and the slave threads position in the master binlog noted. 10 sysbench clients, each configured with 10 threads, were spawned at the same time to generate a random schema load against each of the 10 schemas on the master. Each sysbench client executed 10,000 "update key" statements: UPDATE sbtest set k=k+1 WHERE id = <random row> In total, this generated 100,000 update statements to later replicate during the test itself. Test Methodology: The number of slave workers to test with was configured using: SET GLOBAL slave_parallel_workers=<workers> Then the slave IO thread was started and the test waited for all the update queries to be copied over to the relay log on the slave. The benchmark clock was started and then the slave SQL thread was started. The test waited for the slave SQL thread to finish executing the 100k update queries, doing "select master_pos_wait()". When master_pos_wait() returned, the benchmark clock was stopped and the duration calculated. The calculated duration from the benchmark clock should be close to the time it took for the SQL thread to execute the 100,000 update queries. The 100k queries divided by this duration gave the benchmark metric, reported as Queries Per Second (QPS). Test Reset: The test-reset cycle was implemented as follows: · the slave was stopped · the slave data directory replaced with the previous backup · the slave restarted with the slave threads replication pointer repositioned to the point before the update queries in the binlog. The test could then be repeated with identical set of queries but a different number of slave worker threads, enabling a fair comparison. The Test-Reset cycle was repeated 3 times for 0-24 number of workers and the QPS metric calculated and averaged for each worker count. MySQL Configuration The relevant configuration settings used for MySQL are as follows: binlog-format=STATEMENT relay-log-info-repository=TABLE master-info-repository=TABLE As described in the test procedure, the slave_parallel_workers setting was modified as part of the test logic. The consequence of changing this setting is: 0 worker threads:    - current (i.e. single threaded) sequential mode    - 1 x IO thread and 1 x SQL thread    - SQL thread both reads and executes the events 1 worker thread:    - sequential mode    - 1 x IO thread, 1 x Coordinator SQL thread and 1 x Worker thread    - coordinator reads the event and hands it to the worker who executes 2+ worker threads:    - parallel execution    - 1 x IO thread, 1 x Coordinator SQL thread and 2+ Worker threads    - coordinator reads events and hands them to the workers who execute them Results Figure 1 below shows that Multi-Threaded Slaves deliver ~5x higher replication performance when configured with 10 worker threads, with the load evenly distributed across our 10 x schemas. This result is compared to the current replication implementation which is based on a single SQL thread only (i.e. zero worker threads). Figure 1: 5x Higher Performance with Multi-Threaded Slaves The following figure shows more detailed results, with QPS sampled and reported as the worker threads are incremented. The raw numbers behind this graph are reported in the Appendix section of this post. Figure 2: Detailed Results As the results above show, the configuration does not scale noticably from 5 to 9 worker threads. When configured with 10 worker threads however, scalability increases significantly. The conclusion therefore is that it is desirable to configure the same number of worker threads as schemas. Other conclusions from the results: · Running with 1 worker compared to zero workers just introduces overhead without the benefit of parallel execution. · As expected, having more workers than schemas adds no visible benefit. Aside from what is shown in the results above, testing also demonstrated that the following settings had a very positive effect on slave performance: relay-log-info-repository=TABLE master-info-repository=TABLE For 5+ workers, it was up to 2.3 times as fast to run with TABLE compared to FILE. Conclusion As the results demonstrate, Multi-Threaded Slaves deliver significant performance increases to MySQL replication when handling multiple schemas. This, and the other replication enhancements introduced in MySQL 5.6 are fully available for you to download and evaluate now from the MySQL Developer site (select Development Release tab). You can learn more about MySQL 5.6 from the documentation  Please don’t hesitate to comment on this or other replication blogs with feedback and questions. Appendix – Detailed Results

    Read the article

  • SugarCRM - how to make users see only child users?

    - by John
    Hello, I want to change the behaviour of Sugar CRM community version. Here's what sugar currently does: 1) Admin logs into Sugar. 2) Admin clicks on Admin tab. 3) Admin creates a new user named George with admin access 4) Under user information section, Admin makes George report to Admin (in the database, it will show users.report_to_id is the admin's user_id) 5) Admin saves and logs out 6) George logs in with his password 7) George goes to admin tab. 8) George goes to list users page and sees all users, including Admin, the person he is supposed to report to. I want to change step 8 such that George is not allowed to see the user he reports to. Additionally, he is not allowed to see his grand parent user, great grand parent user, great great grand parent user etc... George should only be able to see users that he creates. How can I achieve this? Is this even possible?

    Read the article

  • Tae Kwon Do in Overland Park

    - by [C.B.W]
    If you are in the Overland Park area and are in need of some physical recreation (and who isn’t) I have to recommend Master’s Tae Kwon Do in Overland Park KS . Master Tom is an 8th Dan teaching Tae Kwon Do and Hapkido. Yah, he teaches almost all of the classes himself. I used to take ishin ryu but stopped some 12 years ago (seems like yesterday. God I am getting old.)    I had wanted to get back into some type of Martial Arts training and I wanted to get my son involved as well – Master’s Tae Kwon Do has the best schedule.   My son and I can go to any of the classes together. Tae Kwon Do is a pretty good work out, lots of kicks so gets the blood pumping. Work out and learn how to defend yourself all at one time. Great for those of us short on time.

    Read the article

  • Crystal Report, Group based Summary at report footer

    - by yesosuresh
    HI, Is it possible to display summary of a group at the report footer ? Let me explain the scenario. I've to create the following kind of report: Client Amount Currency Customer 123: WWWW 300 SGD XXXX 400 SGD YYYY 200 USD ZZZZ 300 USD Customer 456: W1W1 300 SGD X1X1 400 SGD Y1Y1 200 USD Z1Z1 300 USD . . . . . At the report footer, I need to show the following: Grand total: USD 1000 Grand total: SGD 1400 Need to display grand total of amount by Currency at Report footer.Is it possible to do in Crystal report? I'm using crystal report version 11. Could anyone suggest me a solution ?

    Read the article

  • Enterprise Data Quality - New and Improved on Oracle Technology Network

    - by Mala Narasimharajan
    Looking for Enterprise Data Quality technical and developer resources on your projects? Wondering where the best place is to go for finding the latest documentations, downloads and even code samples and libraries?  Check out the new and improved Oracle Technical Network pages for Oracle Enterprise Data Quality.  This section features developer forums as well for EDQ and Master Data Management so that you can connect with other technical professionals who have submitted concerns or posted tips and tricks and learn from them.  Here are the links to bookmark:    Oracle Technology Network website * NEW *   Installation Guide for Enterprise Data Quality Address Verification  Enterprise Data Quality Forum For more information on Oracle's software offerings for data quality and master data management visit:  http://www.oracle.com/us/products/applications/master-data-management/index.html http://www.oracle.com/us/products/middleware/data-integration/enterprise-data-quality/overview/index.html   

    Read the article

  • python: how to convert list of lists into a single nested list

    - by Bhuski
    I have a python list of lists as shown below: mylist=[ [['orphan1', ['some value1']]], [['parent1', ['child1', ['child', ['some value2']]]]], [['parent1', ['child2', ['child', ['some value3']]]]] ] I need to convert the above list to some thing like this: result=[ ['orphan1', ['some value1']], ['parent1', ['child1', ['child', ['some value2']]], ['child2', ['child', ['some value3']]]] ] Kindly help me approach this problem. I have given only simple list. In actual scenario here, in my list, even grand parents/grand childs are there. How much ever deep the input nested list is, I need to convert it to a single nested list, with common list elements (parents and grand parents) appearing only once. (but the next to innermost list element('child' in above example) should appear as many times it occurs in the input list. I have been trying to do this last two days, but did not end up with working solution :(. I need to use the output in django template filter: unordered_list so that the resultant nested list appears as a nested unordered list in my html page ..

    Read the article

  • extrapolating object state based on updates

    - by user494461
    I have a networked multi-user collaborative application. To maintain a consistent virtual world, I send updates for objects from a master peer to a guest peer. The update state contains x,y,z coordinates of object center and his rotation matrix(CHAI3d api used a 3x3 matrix) with 30Hz frequency. I want to reduce this update rate and want to send with a reduced update rate. I want a predictor on both peers. When the predicted value is outside, say a error value of 10% in comparison to master peers objects original state the master peer triggers a state update. Now for position I used velocity,position updates so that the guest peer can extrapolate position. Like velocity for position what parameter should I use for rotation extrapolition?

    Read the article

  • iPhone Development - Query related records using CoreData

    - by Mustafa
    I have a case where i have three entities with one-to-many and one-to-many relationships: Entity A (Entity B relationhip), Entity B (Entity A relationship, Entity C relationship), Entity C (Entity B relationhip) I have the reference of Entity A, and now i want to fetch all the related Entity C records. How can i do that? (with least amount of code) Edit: Here's another way to put it. Can we perform joins with CoreData. For example, (and this is a very crude example), We have a following entity-relationship: Grand Parent (1)---(m) Parent Parent (1)---(m) Child So, now if i have "Albert" the Grand Parent, and i want to get all his grand children, how can i do that?

    Read the article

  • Database IDs

    - by fatherjack
    Just a quick post, mainly to test out the new blog format but related to a question on the #sqlhelp hashtag. The question came from Justin Dearing (@zippy1981) as: So I take it database_id isn’t an ever incrementing value. #sqlhelp When a new database is created it is given the lowest available ID. This either is in a gap in IDs where a database has been dropped or the database ID is incremented by one from the highest current ID if there are no gaps to fill. To see this in action, connect to your sandbox server and try this: USE MASTER GO CREATE DATABASE cherry GO USE cherry GO SELECT DB_ID() GO CREATE DATABASE grape GO USE grape GO SELECT DB_ID() GO CREATE DATABASE melon GO USE melon GO SELECT DB_ID() GO USE MASTER GO DROP DATABASE grape GO CREATE DATABASE kiwi GO USE kiwi GO SELECT DB_ID() GO USE MASTER GO DROP DATABASE cherry DROP DATABASE melon DROP DATABASE kiwi You should get an incrementing series of database IDs as the databases are created until the last one where the new database gets allocated the ID that is missing because one was dropped.

    Read the article

  • Useful git commit messages for merged branches

    - by eykanal
    As a follow-up to this question: If I'm working on a team by myself, I can maintain useful commit messages when merging branches by squashing all the commits to a single diff and then merging that diff. That way I can easily see what changes were introduced in the branch, and I have a single summary describing the feature/change/whatever that was accomplished in that branch when browsing the master branch. My question now is, how can I accomplish this when working with a team? In that situation, the branches will be pushed to a remote repository, meaning that I can't squash all the commits in the branch down to a single commit. If the branch is public, can I still have a single useful merge commit in the master branch? (By "useful" I mean that the commit in the master line tells me (1) a useful summary of what was done in the branch and (2) diffs of the same.)

    Read the article

< Previous Page | 36 37 38 39 40 41 42 43 44 45 46 47  | Next Page >