On 10/11/2011 08:42 AM, Justin Gronfur wrote:
On 10/10/2011 02:46 PM, Rich Megginson wrote:
> I'm just trying to figure out what happens every 5 minutes inside the
> directory server. There is a thread that attempts to clean up
> tombstone entries and other state information. But if you are not
> using replication then that is not run. There are database threads
> such as checkpointing, log flushing, etc. but they run every 250
> milliseconds. That's what this seems like - do you notice high I/O
> usage during the time when the server is pegged at 100% cpu?
No high disk I/O, in fact vmstat is reporting substantially less disk
IO at that time than normal. The cpu is 94% user and 6% system during
those time. I'm not sure if attachments are allowed on this list, but
I used ptrace to grab the list of current operations (not full traces)
and filtered out all of the common select(), poll(), wait() calls.
What is left is attached.
Let me know if you want me to try anything else or gather anymore data.
This is
helpful. Any chance you could paste the entire stack traces?
For example,
#0 0x0000003735c3c868 in slapi_get_mapping_tree_node_by_dn@plt () from
/usr/lib64/dirsrv/libslapd.so.0
#0 0x0000003735c4ad38 in slapi_dn_normalize_ext () from
/usr/lib64/dirsrv/libslapd.so.0
etc. are nice to have, but much better would be the entire stack traces
of these calls so we can see where they are called from.
Thanks,
Justin