ovdb, no more flushing of log.xxxx

[Christiaan den Besten]

Hi all !

We have ovdb running on multiple spool servers, some with bdb-4.2.52 (patched), and some with bdb-4.3.27. After running for several 
hours (sometimes days) ovdb is not able to checkpoint (eh, flush?) its log.xxxxx files to the ovxxxx files. Killing all running 
nnrps (possibly keeping a lock open?) does not work. ovdb_stat -t does show some unfinished transactions ...

Has anyone here who is using ovdb seen this behaviour?

ovdb doesn't seem to be shutting down clean either ...  :
---
news      2059  0.0  0.0  2684  740 ?        S    16:44   0:00 ovdb_monitor
news      2062  0.0  0.0     0    0 ?        Z    16:44   0:00  \_ [ovdb_monitor] <defunct>
news      2061  0.0  4.4 351752 186272 ?     S    16:44   0:02  \_ - ovdb_monitor: checkpoint
---

[root at spool overview_log]# strace -p 2061
Process 2061 attached - interrupt to quit
select(0, NULL, NULL, NULL, {0, 881000}) = 0 (Timeout)
select(0, NULL, NULL, NULL, {1, 0})     = 0 (Timeout)
select(0, NULL, NULL, NULL, {1, 0})     = 0 (Timeout)
select(0, NULL, NULL, NULL, {1, 0})     = 0 (Timeout)
select(0, NULL, NULL, NULL, {1, 0} <unfinished ...>

FYI, DB_CONFIG has the setting:

---
[root at spool root]# cat /var/news/overview_ovdb/DB_CONFIG
set_lg_regionmax        1048576
set_lg_max              10485760
set_lg_bsize            2097152
set_lg_dir              /var/news/overview_log

set_data_dir            /var/news/overview_ovdb

set_lk_max_lockers      40000
set_lk_max_locks        40000
set_lk_max_objects      40000
---

bye,
Chris