lockers and bouncing messages

Andrew Morgan morgan at orst.edu
Mon Feb 13 13:36:19 EST 2006


After upgrading to 2.2.12 this weekend, I'm seeing some sporadic bounce 
messages for incoming mail:

Feb 13 10:12:13 smtp2 postfix/lmtp[29737]: EDC1B1702DF: 
to=<chalainn at onid.orst.edu>, 
relay=cyrus-be1.onid.oregonstate.edu[128.193.4.141], delay=3, 
status=bounced (host cyrus-be1.onid.oregonstate.edu[128.193.4.141] said: 
563 lockers (in reply to end of DATA command))

I'm using the following backends:

annotation_db: skiplist
duplicate_db: berkeley-nosync
mboxlist_db: skiplist
quota_db: quotalegacy
seenstate_db: skiplist
subscription_db: flat
tlscache_db: berkeley-nosync

This is a Debian stable system using libdb4.3 (4.3.27-2). I assume this 
message is coming from the db4.3 based duplicate_db, but I don't 
understand why it is generating a bounce message in the lmtp response. 
The sending mail server is probably interpreting "563" as a SMTP/LMTP 
status code.  :(

Here is what I see for db4 stats:

# db4.3_stat -c -h /var/spool/cyrus/config/db
33224   Last allocated locker ID
0x7fffffff      Current maximum unused locker ID
9       Number of lock modes
50000   Maximum number of locks possible
50000   Maximum number of lockers possible
50000   Maximum number of lock objects possible
291     Number of current locks
309     Maximum number of locks at any one time
582     Number of current lockers
611     Maximum number of lockers at any one time
2       Number of current lock objects
10      Maximum number of lock objects at any one time
2508007 Total number of locks requested
2507716 Total number of locks released
0       Total number of lock requests failing because DB_LOCK_NOWAIT was 
set
370     Total number of locks not immediately available due to conflicts
0       Number of deadlocks
0       Lock timeout value
0       Number of locks that have timed out
0       Transaction timeout value
0       Number of transactions that have timed out
17MB 720KB      The size of the lock region
471     The number of region locks that required waiting (0%)

And yes, the number of lockers does appear to be increasing.

I'm considering switching duplicate and tls_cache to skiplist.

Maybe this email ends up just being another data point against using 
berkeley db (4.3 at least), but if anyone has any comments are 
suggestions, I'd love to hear them.

Thanks,
 	Andy


More information about the Info-cyrus mailing list