[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Attachment: 6170.add2 Addendum 2 addresses Lars F's finding. > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170.add2, 6170.addendum, 6170v4.patch, > HBASE-6170v1.patch, HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Attachment: 6170.addendum Addendum that modifies hbase-default.xml > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170.addendum, 6170v4.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Release Note: "hbase.regionserver.lease.period" is no longer supported. hbase.regionserver.rowlock.timeout.period controls the timeout for rowlock. Default is 60 seconds. hbase.client.scanner.timeout.period controls scanner timeout. Default is 60 seconds. was: hbase.regionserver.rowlock.timeout.period controls the timeout for rowlock. Default is 60 seconds. hbase.client.scanner.timeout.period controls scanner timeout. Default is 60 seconds. Hadoop Flags: Incompatible change,Reviewed (was: Reviewed) > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170v4.patch, HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Release Note: hbase.regionserver.rowlock.timeout.period controls the timeout for rowlock. Default is 60 seconds. hbase.client.scanner.timeout.period controls scanner timeout. Default is 60 seconds. Hadoop Flags: Reviewed > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170v4.patch, HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Attachment: 6170v4.patch Patch v4 wraps a long line (HBASE_REGIONSERVER_ROWLOCK_TIMEOUT_PERIOD). > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170v4.patch, HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Ted Yu updated HBASE-6170: -- Status: Open (was: Patch Available) > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: 6170v4.patch, HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Attachment: HBASE-6170v3.patch @Ted Thanks! Removed HBASE_REGIONSERVER_LEASE_PERIOD_KEY and DEFAULT_HBASE_REGIONSERVER_LEASE_PERIOD. Now the usage should be consistent within the code. Also TestAtomicOperation and TestAccessController tests are passing locally. > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch, HBASE-6170v3.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Attachment: HBASE-6170v2.patch @Ted. Thanks for the reply! I have corrected the comment in HRegionServer. Regarding your second comment, I am not sure I see the need to switch getInt back to getLong. Internally, the timeout value is treated as an int. There is a hidden potential for precision loss if a client uses a long in the config file. Changing everything to getInt removes the cast and makes the internal/external representation of the timeout period consistent. Going forward, if we decide that the extra bits in a long are needed (i.e. a use case that requires timeout periods longer than 24 days), existing configurations will not be broken and the switch to longs will require minimal code changes. > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch, HBASE-6170v2.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Attachment: HBASE-6170v1.patch One more run. Chris > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch, HBASE-6170v1.patch, > HBASE-6170v1.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Attachment: HBASE-6170v1.patch Last run failed due to OOM exception. Resubmitting the patch to get another run. Chris > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch, HBASE-6170v1.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Status: Patch Available (was: Open) Submitting patch for QA run. > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Assignee: Chris Trezzo >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6170) Timeouts for row lock and scan should be separate
[ https://issues.apache.org/jira/browse/HBASE-6170?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Chris Trezzo updated HBASE-6170: Attachment: HBASE-6170v1.patch Attached is a first patch for this issue. I have added two config parameters (and associated defaults) in HConstants: HBASE_REGIONSERVER_ROWLOCK_TIMEOUT_PERIOD HBASE_CLIENT_SCANNER_TIMEOUT_PERIOD Now, each lease in Leases.java can specify their own timeout period. If no timeout period is specified, the original HBASE_REGIONSERVER_LEASE_PERIOD_KEY parameter is used as default. I ran the TestScannerTimeout tests and everything passed. I will submit the patch and get a QA run. Comments are welcome. Thanks, Chris > Timeouts for row lock and scan should be separate > - > > Key: HBASE-6170 > URL: https://issues.apache.org/jira/browse/HBASE-6170 > Project: HBase > Issue Type: Improvement > Components: regionserver >Affects Versions: 0.94.0 >Reporter: Otis Gospodnetic >Priority: Minor > Fix For: 0.96.0 > > Attachments: HBASE-6170v1.patch > > > Apparently the timeout used for row locking and for scanning is global. It > would be better to have two separate timeouts. > (opening the issue to make Lars George happy) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
