[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409664#comment-16409664 ] Vyacheslav Daradur commented on IGNITE-5357: [~dpavlov], thanks for your help. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409652#comment-16409652 ] Dmitriy Pavlov commented on IGNITE-5357: One small change I've done - I've migrated tests fromm cache 5 to cache 6 suite. Cache 6 is not so overloaded as cache 5. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409649#comment-16409649 ] Dmitriy Pavlov commented on IGNITE-5357: Hi [~daradurvs], Merged your changes. Thank you for contribution. Lets continue to keep an eye to TC. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409637#comment-16409637 ] ASF GitHub Bot commented on IGNITE-5357: Github user asfgit closed the pull request at: https://github.com/apache/ignite/pull/3578 > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16409624#comment-16409624 ] Vyacheslav Daradur commented on IGNITE-5357: Here is new [ci.build|https://ci.ignite.apache.org/viewLog.html?buildId=1151311] after review's notest fixes. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16402141#comment-16402141 ] Nikolay Izhikov commented on IGNITE-5357: - I detached abandoned PR link. Now, only actual PR link exists: PR to review - https://github.com/apache/ignite/pull/3578 > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16402129#comment-16402129 ] Dmitriy Pavlov commented on IGNITE-5357: Hi Folks, could you please close not actual PRs. It brings a lot of confusion during review. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16396575#comment-16396575 ] Vyacheslav Daradur commented on IGNITE-5357: [~NIzhikov], thank you so much. Latest ci.build: https://ci.ignite.apache.org/viewLog.html?buildId=1134466 > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16395446#comment-16395446 ] Nikolay Izhikov commented on IGNITE-5357: - [~ascherbakov] I've reviewed update PR one more time. looks good to me. [~daradurvs] Please, provide TC link for current PR changes. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16383455#comment-16383455 ] Nikolay Izhikov commented on IGNITE-5357: - > https://github.com/apache/ignite/pull/3578 Looks good to me. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[
https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16381629#comment-16381629
]
Vyacheslav Daradur commented on IGNITE-5357:
Hi, [~ascherbakov], I've updated [the
PR|https://github.com/apache/ignite/pull/3578/files] to avoid memory's
overhead. Please have a look, if you have time.
Now it covers {{get}} and {{getAll}} operations. I moved logic to
{{GridCacheUtils}} since classes have no suitable common superclasses in the
hierarchy.
New
[ci.tests|https://ci.ignite.apache.org/viewLog.html?buildId=1115559&tab=buildResultsDiv&buildTypeId=IgniteTests24Java8_RunAll]
look good.
About such optimization for {{READ_COMMITED}} mode, I'd like to implement it in
a separate ticket, since it may need significant changes.
> Replicated cache reads load balancing.
> --
>
> Key: IGNITE-5357
> URL: https://issues.apache.org/jira/browse/IGNITE-5357
> Project: Ignite
> Issue Type: Bug
> Components: cache
>Affects Versions: 1.6
>Reporter: Alexei Scherbakov
>Assignee: Lukjanenko Dmitry
>Priority: Major
> Labels: newbie
> Fix For: 2.5
>
>
> Currently all read requests from client node to replicated cache will go
> through primary node for key.
> Need to select random affinity node in topology and send request here (only
> if readFromBackups=true)
> If where are server nodes collocated on same host with client, must select
> target node from them.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16380006#comment-16380006 ] Lukjanenko Dmitry commented on IGNITE-5357: --- There wasn't any agreement with me. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16379960#comment-16379960 ] Vyacheslav Daradur commented on IGNITE-5357: I assigned the ticket to myself in agreement with Dmitry L. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Vyacheslav Daradur >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16379903#comment-16379903 ] Vyacheslav Daradur commented on IGNITE-5357: [~ascherbakov], please have a look at the prepared PR: [https://github.com/apache/ignite/pull/3578/files] [Ci.tests of PR|https://ci.ignite.apache.org/viewLog.html?buildId=1113498&tab=buildResultsDiv&buildTypeId=IgniteTests24Java8_RunAll] look the same as [the master branch|https://ci.ignite.apache.org/viewLog.html?buildId=1113404&tab=buildResultsDiv&buildTypeId=IgniteTests24Java8_RunAll]. Should we add any special test cases for this change, I think it's covered by existing tests? > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Lukjanenko Dmitry >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16379552#comment-16379552 ] ASF GitHub Bot commented on IGNITE-5357: GitHub user DALukjanenko opened a pull request: https://github.com/apache/ignite/pull/3581 IGNITE-5357 replicated cache reads load balancing was added You can merge this pull request into a Git repository by running: $ git pull https://github.com/DALukjanenko/ignite IGNITE-5357.master Alternatively you can review and apply these changes as the patch at: https://github.com/apache/ignite/pull/3581.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #3581 commit 5cc0906838f4916060149a62b51d6904f63a11a7 Author: Dmitry Lukjanenko Date: 2018-02-28T00:14:36Z IGNITE-5357 replicated cache reads load balancing was added > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Lukjanenko Dmitry >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16354586#comment-16354586 ] Lukjanenko Dmitry commented on IGNITE-5357: --- Now it's ok. My bad :) > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Lukjanenko Dmitry >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353912#comment-16353912 ] Lukjanenko Dmitry commented on IGNITE-5357: --- [~ascherbakov], No I don't. I can't see any button to assign. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353735#comment-16353735 ] Alexei Scherbakov commented on IGNITE-5357: --- [~DALukjanenko], Don't you able to reassign it by yourself ? Thanks for taking interest on this. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353689#comment-16353689 ] Lukjanenko Dmitry commented on IGNITE-5357: --- [~ascherbakov], I would like you to assign this issue to me. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353630#comment-16353630 ] Mikhail Lipkovich commented on IGNITE-5357: --- [~DALukjanenko], please go ahead with this issue > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16351524#comment-16351524 ] Lukjanenko Dmitry commented on IGNITE-5357: --- [~mlipkovich], are you going to fix this issue? If you don't mind, I'll take it. > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich >Priority: Major > Labels: newbie > Fix For: 2.5 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16203383#comment-16203383 ] Mikhail Lipkovich commented on IGNITE-5357: --- [~vozerov], [~ascherbakov] thanks both of you! Sorry for late response - I was on a vacation. So probably the question about version is not relevant anymore and it can be changed to 2.4 > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich > Labels: newbie > Fix For: 2.3 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[
https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16186027#comment-16186027
]
Alexei Scherbakov commented on IGNITE-5357:
---
[~mlipkovich],
Yes, you should choose among alive nodes in case of locked topology.
Something like:
{code}
// Always use primary, if reads from backups are not allowed.
if (!cctx.config().isReadFromBackup())
return affNodes.get(0);
Object mac = cctx.localNode().attribute(IgniteNodeAttributes.ATTR_MACS);
assert mac != null;
int r = ThreadLocalRandom8.current().nextInt(affNodes.size());
int c = 0;
ClusterNode n0 = null;
int lastMatch = -1;
for (int i = 0; i < affNodes.size(); i++) {
ClusterNode node = affNodes.get(i);
if (canRemap || cctx.discovery().alive(node)) {
// Prefer collocated node.
if (mac.equals(node.attribute(IgniteNodeAttributes.ATTR_MACS)))
return node;
if (c++ == r)
n0 = node;
lastMatch = i;
}
}
if (n0 == null && lastMatch != -1)
n0 = affNodes.get(lastMatch);
return n0;
{code}
> Replicated cache reads load balancing.
> --
>
> Key: IGNITE-5357
> URL: https://issues.apache.org/jira/browse/IGNITE-5357
> Project: Ignite
> Issue Type: Bug
> Components: cache
>Affects Versions: 1.6
>Reporter: Alexei Scherbakov
>Assignee: Mikhail Lipkovich
> Labels: newbie
> Fix For: 2.3
>
>
> Currently all read requests from client node to replicated cache will go
> through primary node for key.
> Need to select random affinity node in topology and send request here (only
> if readFromBackups=true)
> If where are server nodes collocated on same host with client, must select
> target node from them.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[ https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16185831#comment-16185831 ] Vladimir Ozerov commented on IGNITE-5357: - [~ascherbakov], could you please assist? [~mlipkovich], BTW, there will be a code freeze for AI 2.3 release on Oct 3. Do you think it makes sense to move the ticket to the version 2.4? > Replicated cache reads load balancing. > -- > > Key: IGNITE-5357 > URL: https://issues.apache.org/jira/browse/IGNITE-5357 > Project: Ignite > Issue Type: Bug > Components: cache >Affects Versions: 1.6 >Reporter: Alexei Scherbakov >Assignee: Mikhail Lipkovich > Labels: newbie > Fix For: 2.3 > > > Currently all read requests from client node to replicated cache will go > through primary node for key. > Need to select random affinity node in topology and send request here (only > if readFromBackups=true) > If where are server nodes collocated on same host with client, must select > target node from them. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (IGNITE-5357) Replicated cache reads load balancing.
[
https://issues.apache.org/jira/browse/IGNITE-5357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16174686#comment-16174686
]
Mikhail Lipkovich commented on IGNITE-5357:
---
Hi Alexei,
I'm working on this task but I faced with one question which I can't resolve.
There is a {{canRemap}} flag which has an unclear meaning for me. Here is the
relevant piece of code that I'm going to change according to the task in
{{GridPartitionedSingleGetFuture#affinityNode}}:
{code}
if (!canRemap) {
for (ClusterNode node : affNodes) {
if (cctx.discovery().alive(node))
return node;
}
return null;
}
else
return affNodes.get(0);
{code}
It's commented that {{canRemap}} is a flag indicating that get should be done
on a locked topology version but it's not clear for my how is it related to
piece of code above. Should I somehow take care about this flag during
implementation of requested functionality or I can just ignore it?
Thanks
> Replicated cache reads load balancing.
> --
>
> Key: IGNITE-5357
> URL: https://issues.apache.org/jira/browse/IGNITE-5357
> Project: Ignite
> Issue Type: Bug
> Components: cache
>Affects Versions: 1.6
>Reporter: Alexei Scherbakov
>Assignee: Mikhail Lipkovich
> Labels: newbie
> Fix For: 2.3
>
>
> Currently all read requests from client node to replicated cache will go
> through primary node for key.
> Need to select random affinity node in topology and send request here (only
> if readFromBackups=true)
> If where are server nodes collocated on same host with client, must select
> target node from them.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
