[SVN-2159] Berkeley DB runs out of the log region

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: all
• Fix Version/s: unscheduled
• Component/s: libsvn_fs_base
• Assignee: Unassigned
• Reporter: Justin Erenkrantz
• Created: 09/Dec/04 16:13
• Updated: 27/Oct/18 16:36
• Resolved: 05/Oct/05 16:48

Description

A frequent occurance is starting to be that BDB backends are seeing the following:

db_recover: PANIC: fatal region error detected; run recovery
db_recover: Finding last valid log LSN: file: 2685 offset 819404
db_recover: Recovery starting from [2684][28]
db_recover: Logging region out of memory; you may need to increase its size
db_recover: Recovery function for LSN 2685 707402 failed on backward pass
db_recover: PANIC: Cannot allocate memory
db_recover: PANIC: fatal region error detected; run recovery
...many more fatal region error detected...

svnadmin recover will just have the last error (fatal region error detected; run
recovery).

As described in:

<http://www.contactor.se/~dast/svnusers/archive-2004-10/1001.shtml>

The fix is to modify DB_CONFIG to have:

'set_lg_regionmax 524288'

and rerun svnadmin recover.

C-Mike suggests:

> This should be part of the default DB_CONFIG.  We can go ahead and set
> it to something smallish, 100K or whatever (the lowest you can go is,
> IIRC, 61440) but with a comment about when to up it (the "log region
> out of memory" error) and pointer to BDB docs.