Re: [Nagios-devel] More performance optimizations for 3.2.0,
Posted: Thu Jan 28, 2010 9:36 am
On 01/28/2010 09:49 AM, Ton Voon wrote:
> Hi Jonathan,
>
> On 28 Jan 2010, at 02:28, Kamens, Jonathan wrote:
>
>> See the attached patch, nagios-3.2.0-startup-speed.patch, which
>> optimizes startup time. It requires the patch I sent last month. I've
>> also attached and updated version of that patch which has some bug
>> fixes in it. Enjoy!
>
> Thanks. I've noticed this too for comments and downtimes.
>
> I'm wondering if there is a better way of doing this. Rather than each
> caller having to work out the next comment id value and making sure it
> is still unique, I think the next_comment_id should be manipulated
> within xrddefault.c instead. If you call add_new_comment with a
> comment_id of 0, it uses next_comment_id and ensures it gets incremented.
>
> I guess on the initial load of retention.dat you have to validate that
> the next_comment_id is unique, but you shouldn't have to after that.
>
> Can you extend the tests in t-tap/ to time the performance increase?
>
> So I'm thinking that the patch needs a bit more work to be applied.
>
Seconded. The API should take care of making sure comment id's are
unique and incremental. In fact, passing a comment id to the API at
all is, to me, totally superfluous and counterproductive.
--
Andreas Ericsson [email protected]
OP5 AB www.op5.se
Tel: +46 8-230225 Fax: +46 8-230231
Considering the successes of the wars on alcohol, poverty, drugs and
terror, I think we should give some serious thought to declaring war
on peace.
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]
> Hi Jonathan,
>
> On 28 Jan 2010, at 02:28, Kamens, Jonathan wrote:
>
>> See the attached patch, nagios-3.2.0-startup-speed.patch, which
>> optimizes startup time. It requires the patch I sent last month. I've
>> also attached and updated version of that patch which has some bug
>> fixes in it. Enjoy!
>
> Thanks. I've noticed this too for comments and downtimes.
>
> I'm wondering if there is a better way of doing this. Rather than each
> caller having to work out the next comment id value and making sure it
> is still unique, I think the next_comment_id should be manipulated
> within xrddefault.c instead. If you call add_new_comment with a
> comment_id of 0, it uses next_comment_id and ensures it gets incremented.
>
> I guess on the initial load of retention.dat you have to validate that
> the next_comment_id is unique, but you shouldn't have to after that.
>
> Can you extend the tests in t-tap/ to time the performance increase?
>
> So I'm thinking that the patch needs a bit more work to be applied.
>
Seconded. The API should take care of making sure comment id's are
unique and incremental. In fact, passing a comment id to the API at
all is, to me, totally superfluous and counterproductive.
--
Andreas Ericsson [email protected]
OP5 AB www.op5.se
Tel: +46 8-230225 Fax: +46 8-230231
Considering the successes of the wars on alcohol, poverty, drugs and
terror, I think we should give some serious thought to declaring war
on peace.
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]