Hi Magin,
On 3 Nov 2009, at 02:54, MAGIN GEORGE wrote:
I have a problem with import_logsd . The number of files in /usr/
local/nagios/var/ndologs grows with time. Following is the opsview
setup.
6 slaves
9500 service checks
300 hosts.
As I understand "rotate_ndo_log" executes every 5 seconds which
rotates the file based on timestamp. and import_ndologsd pickup
these files one by one and writes to "/usr/local/nagios/var/
ndo.sock" from where ndo2db picks up and writes to the runtime DB.
So in my understanding file rotation is suppose to happen only once
in 5 secs but in case it happens every second which creates lots of
files in var/ndologs . I think import_ndologsd lags behind due to
the large number of files it has to process.
Co-incidentally, I am just updating some of the documentation around
this area: http://docs.opsview.org/doku.php?id=opsview-community:runtimedb
I guess you are on Opsview 2.14. This rotated the logs more than once
every 5 seconds and also if a host state returned with a non-up state.
We patched this in Opsview 2.14 because Nagios 2.X blocked during host
retries.
We have a another similar master (which is having more number of
servicechecks 10k+ ) but works fine( It rotates file onece in 5 secs
and mostly the file count is 0).
Any clue why nagios is rotating ndo_logs too often ???
My guess is that this other master has less host failures?
Ton
_______________________________________________
Opsview-users mailing list
[email protected]
http://lists.opsview.org/lists/listinfo/opsview-users
