[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

[URGENT] DB_RECOVER doesnt't recover

From: <allan_at_muly.dk>
Date: 2006-02-16 13:10:21 CET

hello

we have (i believe) a pretty serious problem with our repository.

Subversion 1.2
apache 2.0.54

i have done the suggestings with regard to "svnadmin recover"

first i tried

C:\>svnadmin recover D:\path\to\repos\
Repository lock acquired.
Please wait; recovering the repository may take some time...

Recovery completed.
The latest repos revision is 39806.

then restarted apache, but got this in the erro.log:

error.log:
[client 127.0.0.1] Could not fetch resource information. [301, #0],
referer: http://localhost/
[client 127.0.0.1] Requests for a collection must have a trailing slash
on the URI. [301, #0], referer: http://localhost/
[client 127.0.0.1] Unable to deliver content. [500, #0], referer:
http://localhost/
[client 127.0.0.1] could not fetch directory entries [500, #160029],
referer: http://localhost/
[client 127.0.0.1] Berkeley DB error for filesystem D:/path/to/repos/db
while moving cursor:\nDB_RUNRECOVERY: Fatal error, run database
recovery [500, #160029], referer: http://localhost/
[client 127.0.0.1] (20014)Error string not specified yet: Berkeley DB
error for filesystem D:/path/to/repos/db while closing 'nodes'
database:\nDB_RUNRECOVERY: Fatal error, run database recovery, referer:
http://localhost/

then i stop apache

then i downloaded this:
http://subversion.tigris.org/files/documents/15/24470/db-4.3.28-win32.zip

then i ran this C:\>db_recover.exe -c -v -h D:\path\to\repos\db

which printed this:
Finding last valid log LSN: file: 9780 offset 412963
db_recover: Skipping log file D:\path\to\repos\db\log.0000002226:
historic log version 8
db_recover: Skipping log file D:\path\to\repos\db\log.0000001989:
historic log version 8
db_recover: Skipping log file D:\path\to\repos\db\log.0000001867:
historic log version 8
db_recover: Skipping log file D:\path\to\repos\db\log.0000001850:
historic log version 8
db_recover: Skipping log file D:\path\to\repos\db\log.0000001747:
historic log version 8
db_recover: Log file corrupt at LSN: [2845][1045838]
db_recover: PANIC: Invalid argument
db_recover: PANIC: fatal region error detected; run recovery
db_recover: DB_ENV->open: DB_RUNRECOVERY: Fatal error, run database recovery

i'm slighly puzzled and haven't been able to pinpoint this.
i do know that the "corruption" occured after a user tiried to uplaod a
semi-large file (50MB)

BTW i also ran this
C:\>svnadmin lstxns D:\path\to\repos\

which produces about 20-30 messages about uncommited transactions as
well as 3 that was dead
when trying to delete these i get the same type of error message (fatal
region error detected; run recovery)

so what to do, anyone been here before - we are quite desperate since
we have a pretty active repository?

should we upgrade ?
should we move to unix ?

any help veru much appriciated

thanks ./allan

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Thu Feb 16 13:14:28 2006

This is an archived mail posted to the Subversion Users mailing list.