Hi Ken,
Thanks for the follow-up. I will read up more on Sequoia-905.
Meanwhile, it turns out that the commercial case I'm working on is
fixed by one of my own previous fixes to Sequoia (sequoia-974) from
September 4th. Prior to that fix it was fairly easy for recovery
logs to become silently messed up if the database holding them went
away. Maybe it's time for a vacation. It's certainly time for a
new Sequoia build that fixes some of the annoyances that are piling
up like the dumb driver warnings pointed out in a recent post by
Stefan Lischke.
Speaking of fixing things, I didn't get much response to my grand
post on the future of Sequoia. If you are interested in helping to
get things rolling forward again please speak up.
Cheers, Robert
Robert Hodges, CTO, Continuent, Inc.
Email: [EMAIL PROTECTED]
Mobile: +1-510-501-3728 Skype: hodgesrm
On Oct 30, 2007, at 6:37 PM, Ken Burns wrote:
Hi Robert and Dominic,
Reading your posting this morning, it looks like you have a single
controller in your configuration? I had the same issue at the
beginning of
the year, take a look at JIRA issue SEQUOIA-905 - and the email
chain on the
list around the last week of January this year.
Hope this helps,
Regards
Ken
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
Robert
Hodges (JIRA)
Sent: Wednesday, 31 October 2007 8:15 AM
To: [email protected]
Subject: [Sequoia] [JIRA] Commented: (SEQUOIA-996) Recovery log is not
applied when backend is enabled
[
https://forge.continuent.org/jira/browse/SEQUOIA-996?
page=comments#action_14
112 ]
Robert Hodges commented on SEQUOIA-996:
---------------------------------------
Hi Dominic!
There's definitely something not right here...
I'm tracking down a bug from our commercial products that looks really
similar. Where were you keeping the recovery logs? What is
happening in
that particular case is that there's a corner case (or maybe a race
condition) that is causing the controller not to synchronize
logs. It
seems to occur when the controller logs are in the same database as
the VDB
backend.
Would you be so kind as to attach your VDB definition files to the
case?
You can wipe out passwords if that's an issue. I just need to see the
overall configuration, especially of the recovery log. Also, when
you look
in the logs, do you see any errors from the recovery log? If so,
please
append that log as well.
Thanks, Robert
Recovery log is not applied when backend is enabled
---------------------------------------------------
Key: SEQUOIA-996
URL: https://forge.continuent.org/jira/browse/SEQUOIA-996
Project: Sequoia
Type: Bug
Components: Backup System
Versions: Sequoia 2.9
Environment: Linux Redhat with postgres 8.2.5
Reporter: Dominic Vieira
I have setup sequoia with two postgres backends. I activate the
cluster accordling to the user guide documentation. 1. Initialize BE1
2. Enable BE1 3.Create a backup of BE1
4.Restore to BE2
5. Enable BE2
6. Insert 2 rows, they appear in both
7. I disable BE2
8. Insert row 3, appears only in BE1 (good)
9.Enable BE2 **I expected to see the row replicated over to BE2,
but it
doesn't and the backends are out of sync with both backends showing
enabled***
The checkpoints appear correct in the recovery logs and also in "show
backend *"
Is there something wrong here?
--
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
_______________________________________________
Sequoia mailing list
[email protected]
https://forge.continuent.org/mailman/listinfo/sequoia
_______________________________________________
Sequoia mailing list
[email protected]
https://forge.continuent.org/mailman/listinfo/sequoia
