[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Mahadev konar updated ZOOKEEPER-790: Status: Resolved (was: Patch Available) Resolution: Fixed I just committed this. thanks flavio and others! Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.v2.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Status: Open (was: Patch Available) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.v2.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Attachment: ZOOKEEPER-790-3.3.v2.patch Fixed issue pointed out by Mahadev. Submitting patch for 3.3 first. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.v2.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Attachment: ZOOKEEPER-790.v2.patch Now uploading trunk version. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.v2.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Status: Patch Available (was: Open) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.v2.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Sergey Doroshenko updated ZOOKEEPER-790: Attachment: ZOOKEEPER-790-test.patch The test that was failing was the test for r/o mode, so it wouldn't apply here. I've extracted necessary code from it and made it a test case in QuorumTest. It fails against current trunk but passes against latest patch. This test case uses QuorumUtil which I also extracted from my r/o mode patch. I made it for convenient quorum testing, it's more handy than current QuorumBase, and I plan to make QuorumBase a special case of it. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Status: Open (was: Patch Available) Submitting a new patch right after. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Attachment: ZOOKEEPER-790.v2.patch Thanks for providing a code base for the test, Sergey. I liked QuorumUtil so much that I have reimplemented the previous test to use it. This is in fact aligned with a discussion Ben and I had the other day about refactoring QuorumBase. You have introduced pretty much all functionality I was looking for, which simplified the test code. I hope you don't mind that I have made some modifications to your original test. My proposed modifications are to make sure that we induce the right set of events. In the modified version, I have made sure that we have a client connecting to a follower and that we kill the leader so that we force a new election. In fact, the test was not always reliable for me in its original form. The patch works for me. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Status: Patch Available (was: Open) Need to generate a version for 3.3, but the one uploaded is ready to be reviewed for me. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Benjamin Reed updated ZOOKEEPER-790: +1 excellent work you guys. i also like QuorumUtil sergey! thanx for implementing it. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790-test.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Sergey Doroshenko updated ZOOKEEPER-790: Attachment: ZOOKEEPER-790-follower-request-NPE.log It seems this patch introduces a bug: followers get synchronized with leader and start serving clients before leader startups its own zk instance. As the result, when a follower forward request to the leader, for example session revalidation request, it fails because leader's sessionTracker is null. Previously this was not the case because leader started its zk immediately after quorum peer is switched to LEADING state. See attached log. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Attachment: ZOOKEEPER-790.v2.patch Thanks for pointing this issue out, Sergei. It sounds like the previous patch solved the issue discussed without making sure that the leader was ready to process messages when learner handlers started to read them in. This v2 patch does a number of things: # It moves the startup method to processAck. This way we make sure that start up the leader as soon as we have a quorum of acks for the newleader message; # It moves the initialization of the database in startup to a method startdata. There are two reasons for doing it. First, it didn't sound like a good idea to throw exceptions or catch exceptions in processAck, and they were only necessary because of the call to startup(). Second, the method startup() in ZooKeeperServer throws these exceptions because of loadData(), which is called separately in Leader.lead(), so it is not necessary to call it in processAck after hearing from a quorum; # It waits in LearnerHandler.run() until the leader ready before it starts the while(true) loop. I also had to receive an ack before executing the code to wait, otherwise the leader would never receive acks and form a quorum, thus causing the system to halt. To get some feedback on the changes implemented in this patch, I have discussed them with Ben. Thanks, Ben! Sergei, I would appreciate if you could give it a try, and if you could tell me if it works for you. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Junqueira updated ZOOKEEPER-790: --- Status: Patch Available (was: Reopened) The latest patch applies to both trunk and 3.3. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Junqueira Assignee: Flavio Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-follower-request-NPE.log, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2, ZOOKEEPER-790.v2.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Benjamin Reed updated ZOOKEEPER-790: Status: Resolved (was: Patch Available) Resolution: Fixed Committed revision 966960. Committed revision 966984. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Attachment: ZOOKEEPER-790-3.3.patch Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Status: Open (was: Patch Available) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Status: Patch Available (was: Open) Thank for reviewing, Ben. Resubmitting with the proposed fix. Thanks for testing the patch, Travis. I don't really know why you're observing the wrong dates, though. Anyone else? Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Benjamin Reed updated ZOOKEEPER-790: Hadoop Flags: [Reviewed] +1 great job flavio! thanx for your help travis and vishal. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Status: Open (was: Patch Available) Just in case will cancel and resubmit. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Attachment: ZOOKEEPER-790.patch I realized that while working on the patch I removed a private modifier and forgot to add it back for the trunk patch, so I'm uploading an new patch. There is no other significant modification to the patch. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Status: Patch Available (was: Open) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Attachment: ZOOKEEPER-790.patch This patch is very simple: it moves two function calls such that a leader only starts up and sets the the last processed zxid after it has a quorum of supporters. It also includes a unit test. I had to make a few modifications here and there to be able to write this test. I tried to minimize changes as much as possible. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Status: Patch Available (was: Open) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Attachment: ZOOKEEPER-790-3.3.patch Uploading patch for the 3.3 branch. I have also checked the results of Hudson, and I couldn't find any java test failure. The -1 on core tests seems to be unrelated to this patch. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790-3.3.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Flavio Paiva Junqueira updated ZOOKEEPER-790: - Attachment: ZOOKEEPER-790.patch Hi! I'm attaching a patch that works for my setup (and passes the unit tests as well), so I would appreciate if you could give it a try. Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790.patch, ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Travis Crawford updated ZOOKEEPER-790: -- Attachment: ZOOKEEPER-790.travis.log.bz2 Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790.patch, ZOOKEEPER-790.travis.log.bz2 The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Updated: (ZOOKEEPER-790) Last processed zxid set prematurely while establishing leadership
[ https://issues.apache.org/jira/browse/ZOOKEEPER-790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Patrick Hunt updated ZOOKEEPER-790: --- Assignee: Flavio Paiva Junqueira Fix Version/s: 3.4.0 Priority: Blocker (was: Major) Last processed zxid set prematurely while establishing leadership - Key: ZOOKEEPER-790 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-790 Project: Zookeeper Issue Type: Bug Components: quorum Affects Versions: 3.3.1 Reporter: Flavio Paiva Junqueira Assignee: Flavio Paiva Junqueira Priority: Blocker Fix For: 3.3.2, 3.4.0 Attachments: ZOOKEEPER-790.patch The leader code is setting the last processed zxid to the first of the new epoch even before connecting to a quorum of followers. Because the leader code sets this value before connecting to a quorum of followers (Leader.java:281) and the follower code throws an IOException (Follower.java:73) if the leader epoch is smaller, we have that when the false leader drops leadership and becomes a follower, it finds a smaller epoch and kills itself. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
