[ 
https://forge.continuent.org/jira/browse/SEQUOIA-419?page=comments#action_13772 
] 

Olivier Fambon commented on SEQUOIA-419:
----------------------------------------

This is (partly) soved by the following commit (RecoveryLog.java 1.110.2.9->11):
https://forge.continuent.org/pipermail/sequoia-commits/2007-February/003724.html
 (code fix)
https://forge.continuent.org/pipermail/sequoia-commits/2007-February/003732.html
 (updated javadoc)

purge now clears the log up to specified checkpoint EXCLUDED.
purge now also clears old checkpoints.

wa(user) > debug on
Debug mode on
wa(user) > show checkpoints
+----+-------------------------------------------------------------------------------------------------------+
| id | name                               |
+----+-------------------------------------------------------------------------------------------------------+
| 0  | Initial_empty_recovery_log                               |
| 0  | Member(address=node50/192.168.0.150:6523, uid=wa) joined group 
wa-node50:25322-20070226190623223+0100 |
| 0  | Member(address=node50/192.168.0.150:6523, uid=wa) joined group 
wa-node50:25322-20070227101510569+0100 |
| 0  | backup wa_Dump_2-node50:25322-20070226190458405+0100                     
          |
| 0  | backup wa_Dump_3-node50:25322-20070226190530452+0100                     
          |
| 2  | backup wa_Dump_4-node50:25322-20070301143040317+0100                     
          |
| 0  | disable all backends-node50:25322-20070226190602195+0100                 
              |
| 0  | disable all backends-node50:25322-20070227101420224+0100                 
              |
| 0  | now-node51:25322-20070226190625027+0100                               |
| 0  | now-node51:25322-20070227101512735+0100                               |
| 0  | shutdown-node50:25322-20070226190602316+0100                             
  |
| 0  | shutdown-node50:25322-20070227101420471+0100                             
  |
+----+-------------------------------------------------------------------------------------------------------+

wa(user) > purge log wa_Dump_4
wa(user) > show checkpoints
+----+------------------------------------------------------+
| id | name                                                 |
+----+------------------------------------------------------+
| 2  | backup wa_Dump_4-node50:25322-20070301143040317+0100 |
+----+------------------------------------------------------+


> "purge log" Command Does Not Remove Associated Checkpoint(s)
> ------------------------------------------------------------
>
>          Key: SEQUOIA-419
>          URL: https://forge.continuent.org/jira/browse/SEQUOIA-419
>      Project: Sequoia
>         Type: Bug
>   Components: Recovery Log
>     Versions: Sequoia 2.8.1
>  Environment: 1 Sequoia Controller (from CVS, checkout March 29/2006, 08:30 
> PDT), 1 virtual database, 1 PostgreSQL 8.1 backend
>     Reporter: Dylan Hansen
>     Assignee: Jeff Mesnil
>      Fix For: Sequoia 3.0 beta3

>
>
> Quoting an e-mail that was sent to the mailing list from Emmanuel:
> "You should never remove checkpoints. The checkpoints are hidden to the user. 
> They will be automatically dropped when you remove a dump or resync/purge the 
> recovery log."
> I have simple setup that I've performed 3 backups on (live backend), which 
> has produced some checkpoints:
> testing(admin) > show dumps 
> +---------+---------------------------------------------+---------------------+-------------------------------------+------------------+---------------------+--------+
> | Name    | Checkpoint                                  | Format              
> | Path                                | Date             | Backend            
>  | Tables | 
> +---------+---------------------------------------------+---------------------+-------------------------------------+------------------+---------------------+--------+
> | backup3 | cp_for_backup3_Wed Mar 29 10:24:25 PST 2006 | PostgreSQL Tar Dump 
> | /usr/local/sequoia/backups/testing/ | 3/29/06 10:24 AM | database1 | *      
> | 
> | backup2 | cp_for_backup2_Wed Mar 29 10:24:18 PST 2006 | PostgreSQL Tar Dump 
> | /usr/local/sequoia/backups/testing/ | 3/29/06 10:24 AM | database1 | *      
> | 
> | backup1 | cp_for_backup1_Wed Mar 29 10:24:08 PST 2006 | PostgreSQL Tar Dump 
> | /usr/local/sequoia/backups/testing/ | 3/29/06 10:24 AM | database1 | *      
> | 
> +---------+---------------------------------------------+---------------------+-------------------------------------+------------------+---------------------+--------+
> testing(admin) > show checkpoints 
> Checkpoints...
> [0]     cp_for_backup1_Wed Mar 29 10:24:08 PST 2006
> [1]     cp_for_backup2_Wed Mar 29 10:24:18 PST 2006
> [2]     cp_for_backup3_Wed Mar 29 10:24:25 PST 2006
> [3]     Initial_empty_recovery_log
> Lets say that I want to purge the log up to backup3.  What I'm assuming this 
> should do is remove all checkpoints previous to the checkpoint that is 
> assoicated with the backup we're purging the log to.  However, this is not 
> the case:
> testing(admin) > purge log backup3
> testing(admin) > show checkpoints 
> Checkpoints...
> [0]     cp_for_backup1_Wed Mar 29 10:24:08 PST 2006
> [1]     cp_for_backup2_Wed Mar 29 10:24:18 PST 2006
> [2]     cp_for_backup3_Wed Mar 29 10:24:25 PST 2006
> [3]     Initial_empty_recovery_log
> The log entries are removed, however the checkpoints still remain.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   https://forge.continuent.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira


_______________________________________________
Sequoia mailing list
[email protected]
https://forge.continuent.org/mailman/listinfo/sequoia

Reply via email to