|
This is not a question, rather I wanted to bring attention to some issues we are seeing with database recovery times in Sybase 12.0.1.3750 Below is a snippet of the recovery log file and the output of dblog. which indicates a log file with 5434 4K pages took > 30 days to apply to a back up database. The hardware is not the issue here. Regards, Jim SQL Anywhere Network Server Version 12.0.1.3750 I. 05/03 10:27:27. Database recovery in progress I. 05/03 10:27:27. Last checkpoint at Sat May 03 2014 10:27 I. 05/03 10:27:27. Transaction log: D:\Database Recovery\Production\Database Log Files Temp\140502AD.LOG... I. 05/03 10:27:33. ... recovery 0% complete ~~~ I. 05/03 16:58:44. ... recovery 60% complete I. 05/11 03:53:37. ... recovery 61% complete I. 05/23 23:23:48. ... recovery 62% complete I. 05/30 14:52:20. ... recovery 63% complete I. 06/05 17:25:34. ... recovery 66% complete I. 06/05 17:25:40. ... recovery 72% complete ~~~ I. 06/05 17:27:00. ... recovery 100% complete I. 06/05 17:27:09. Starting checkpoint of "Production" (Production.db) at Thu Jun 05 2014 17:27 I. 06/05 17:27:45. Finished checkpoint of "Production" (Production.db) at Thu Jun 05 2014 17:27 I. 06/05 17:27:46. Checkpointing... I. 06/05 17:27:46. Starting checkpoint of "Production" (Production.db) at Thu Jun 05 2014 17:27 I. 06/05 17:27:49. Finished checkpoint of "Production" (Production.db) at Thu Jun 05 2014 17:27 I. 06/05 17:27:49. Recovery complete I. 06/05 17:28:01. Database server shutdown automatically after log applied I. 06/05 17:28:02. Database server stopped at Thu Jun 05 2014 17:28 |
Please be aware that the SAP SQL Anywhere Forum will be shut down on August 29th, 2024 when all it's content will be migrated to the SAP Community.
Follow this question
By Email:Once you sign in you will be able to subscribe for any updates here
By RSS:Markdown Basics
- *italic* or _italic_
- **bold** or __bold__
- link:[text](http://url.com/ "title")
- image?
- numbered list: 1. Foo 2. Bar
- to add a line break simply add two spaces to where you would like the new line to be.
- basic HTML tags are also supported
Question tags:
question asked: 10 Jun '14, 08:36
question was seen: 3,143 times
last updated: 14 Jun '14, 15:16
Is the recovery done on the same machine (and with identical command line settings) as the original database?
A similar question with very helpful information by John has been asked a while ago:
Sybase 12 Recovery Times
...I guess you are aware of that:)
Clearly you are recovering with the -a switch. What else is important to know from how you are running this? (including server switches, hardware info. and other information especially if different from production)
What event caused you recover from backups? Or is this part of your validation process?
Are you using some special mode in production such as database mirroring on this system?
If you had the time, one option that is available to you to investigate this deeper yourself, is to translate your 22MB of transaction log and apply that using dbisql and see where the time is takes. That is not strictly the same thing as a recovery but it may help identify what operations are taking so long time intervals incurred between the 60% and 66% times.
Nothing special in production, no mirroring and yes this was run during our validation process. Hardware is a bit old but should suffice Dell PowerEdge 2900 4 GB Memory and 5 TB RAID 5 Drive Space. We use this server to validate incremental backup files as they are created on 9 production databases. It has to be something to do with the operations in the log file because others of equal size apply in a reasonable amount of time. I will do as you suggested and translate the log file and take a look.
Thanks,
Jim