[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16497096#comment-16497096 ] ASF subversion and git services commented on SOLR-12388: Commit 61a65d5928f4563f5c8284b1100528257e5c6834 in lucene-solr's branch refs/heads/branch_7x from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=61a65d5 ] SOLR-12388: Add an expected exception message to SearchHandlerTest.testRequireZkConnectedDistrib() > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) > Components: search, SolrCloud >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Fix For: 7.4, master (8.0) > > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16497097#comment-16497097 ] ASF subversion and git services commented on SOLR-12388: Commit 252a8145d9374978776f3fc10bfa12e14bf8433a in lucene-solr's branch refs/heads/master from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=252a814 ] SOLR-12388: Add an expected exception message to SearchHandlerTest.testRequireZkConnectedDistrib() > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) > Components: search, SolrCloud >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Fix For: 7.4, master (8.0) > > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16497095#comment-16497095 ] ASF subversion and git services commented on SOLR-12388: Commit 114461cbeb2c7a2c9f610a46d4e01ca2ee9cf171 in lucene-solr's branch refs/heads/branch_7x from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=114461c ] SOLR-12388: print out exception when assert fails > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) > Components: search, SolrCloud >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Fix For: 7.4, master (8.0) > > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16494680#comment-16494680 ] ASF subversion and git services commented on SOLR-12388: Commit bd14de24e9515b00198c42107486c99722ed8649 in lucene-solr's branch refs/heads/master from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=bd14de2 ] SOLR-12388: print out exception when assert fails > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) > Components: search, SolrCloud >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Fix For: 7.4, master (8.0) > > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16491845#comment-16491845 ] ASF subversion and git services commented on SOLR-12388: Commit 3c6fb974aefc50f411cba90b2b267804fbb6e1cb in lucene-solr's branch refs/heads/branch_7x from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=3c6fb97 ] SOLR-12388: Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node cant communicate with ZooKeeper, by setting the "shards.tolerant" param to "requireZkConnected" > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16490242#comment-16490242 ] ASF subversion and git services commented on SOLR-12388: Commit 54a63d0d0c250a4aa45a3e3a770efffc106f8ce2 in lucene-solr's branch refs/heads/master from [~steve_rowe] [ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=54a63d0 ] SOLR-12388: Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node cant communicate with ZooKeeper, by setting the "shards.tolerant" param to "requireZkConnected" > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16489480#comment-16489480 ] Steve Rowe commented on SOLR-12388: --- I plan on committing later today if there are no objections. > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16487561#comment-16487561 ] Steve Rowe commented on SOLR-12388: --- bq. I think Steve refers to cluster changes that can happen and a node might have missed out on hearing about. Right. Solr already handles these conditions, as mentioned in the description, via the {{zkConnected}} header. This issue just enables callers to get a failure response instead of having to conditionally handle responses based on the value of the {{zkConnected}} header. > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16487471#comment-16487471 ] Mark Miller commented on SOLR-12388: We stop accepting document updates when we realize we lost the connection to ZK, but I think Steve refers to cluster changes that can happen and a node might have missed out on hearing about. > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16487208#comment-16487208 ] Shawn Heisey commented on SOLR-12388: - Interesting. It's my understanding that SolrCloud goes read-only when ZK quorum is lost, so it would have to be a particularly unusual network partition for the described situation to arise. But as noted by the author of Jepsen, unusual network partitions DO happen in the wild. > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (SOLR-12388) Enable a strict ZooKeeper-connected search request mode, in which search requests will fail when the coordinating node can't communicate with ZooKeeper
[ https://issues.apache.org/jira/browse/SOLR-12388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16487071#comment-16487071 ] Steve Rowe commented on SOLR-12388: --- Attached patch implements the idea. Rather than introducing a new request param, I've expanded the possible values {{shards.tolerant}} can take on to include non-boolean value {{requireZkConnected}}, which enables the mode described above. (Thanks to [~hossman] for his offline suggestion to use {{shards.tolerant}} for this purpose.) In addition to causing requests to fail when the coordinating node can't communicate with ZooKeeper, setting {{shards.tolerant}} to {{requireZkConnected}} will cause search components to behave the same as when {{shards.tolerant}} is set to {{false}} (the default): the request will fail rather than causing partial results to be returned. I've included ref guide docs and a CHANGES entry. Precommit and all Solr tests pass. I think this is ready to go. Feedback is welcome. > Enable a strict ZooKeeper-connected search request mode, in which search > requests will fail when the coordinating node can't communicate with ZooKeeper > --- > > Key: SOLR-12388 > URL: https://issues.apache.org/jira/browse/SOLR-12388 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) >Reporter: Steve Rowe >Assignee: Steve Rowe >Priority: Minor > Attachments: SOLR-12388.patch > > > Right now, a Solr node will return the results of a search request even if it > cannot communicate with ZooKeeper at the time it receives the request. This > may result in stale or incorrect results if there have been major changes to > the collection structure that the node has not been informed of via > ZooKeeper. When this happens, as long as all known shards respond, the > response will succeed, and a {{zkConnected}} header set to {{false}} is > included in the search response. > There should be an option to instead fail requests under these conditions, to > prevent stale or incorrect results. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org