Closed infraweavers closed 4 years ago
Good morning,
We have several OMD checks that return huge amounts of perfdata (~ 80kB) we've noticed that these are completely fine into thruk and naemon in general, however our pnp_gearman_worker seems to receive a truncated set of perfdata.
We've traced this back to here: https://github.com/sni/mod_gearman/blob/6305296842a7fbfbc1aa26602a29322216d492b8/neb_module/mod_gearman.c#L1566
It looks like there's a 64KiB limit on the message size that's enqueued onto the perfdata queue. Is there any particular reason that this is smaller than GM_MAX_OUTPUT (https://github.com/sni/mod_gearman/blob/e56e94be6d677aed55aa776e977a04ffee122cf0/include/common.h#L63)?
perfdata
GM_MAX_OUTPUT
Would you accept a PR changing GM_BUFFERSIZE up to 10485760 ?
GM_BUFFERSIZE
10485760
There is no reason for this, it was a arbitrary limit which i thought it would be enough.
Good morning,
We have several OMD checks that return huge amounts of perfdata (~ 80kB) we've noticed that these are completely fine into thruk and naemon in general, however our pnp_gearman_worker seems to receive a truncated set of perfdata.
We've traced this back to here: https://github.com/sni/mod_gearman/blob/6305296842a7fbfbc1aa26602a29322216d492b8/neb_module/mod_gearman.c#L1566
It looks like there's a 64KiB limit on the message size that's enqueued onto the
perfdata
queue. Is there any particular reason that this is smaller thanGM_MAX_OUTPUT
(https://github.com/sni/mod_gearman/blob/e56e94be6d677aed55aa776e977a04ffee122cf0/include/common.h#L63)?Would you accept a PR changing
GM_BUFFERSIZE
up to10485760
?