osdir.com


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ops][nova][ceilometer] Quick show of hands: any use Intel (non-CMT) `perf` events?


Sorry for reply late.

+ ceilometer

The ceilometer is still using the cmt meter
https://github.com/openstack/ceilometer/blob/master/ceilometer/compute/virt/libvirt/inspector.py#L224
should
we deprecate them?

And there are some other meter depend on perf feature
https://github.com/openstack/ceilometer/blob/master/ceilometer/compute/virt/libvirt/inspector.py#L218

So sounds like we shouldn't remove the whole perf feature.

Thanks
Alex

Kashyap Chamarthy <kchamart at redhat.com> äº?2019å¹´7æ??4æ?¥å?¨å?? ä¸?å??6:37å??é??ï¼?

> Heya folks,
>
> While removing some dead code I was wondering if anyone here uses
> "non-CMT" (Cache Monitoring Technology) performance events events?  I'm
> referring to the events here[0], besides the first three, which are
> CMT-related.
>
> Background
> ----------
>
> The Intel CMT events (there are three of them) were deprecated during
> the Rocky release, in this[1] commit, and with this rationale:
>
>     Upstream Linux kernel has deleted[*] the `perf` framework integration
>     with Intel CMT (Cache Monitoring Technology; or "CQM" in Linux kernel
>     parlance), because the feature was broken by design -- an
>     incompatibility between Linux's `perf` infrastructure and Intel CMT
>     hardware support.  It was removed in upstream kernel version v4.14; but
>     bear in mind that downstream Linux distributions with lower kernel
>     versions than 4.14 have backported the said change.
>
>     Nova supports monitoring of the above mentioned Intel CMT events
>     (namely: 'cmt', 'mbm_local', and 'mbm_total') via the configuration
>     attribute `[libvirt]/enabled_perf_events`. Given that the underlying
>     Linux kernel infrastructure for Intel CMT is removed, we should remove
>     support for it in Nova too.  Otherwise enabling them in Nova, and
>     updating to a Linux kernel 4.14 (or above) will result in instances
>     failing to boot.
>
>     To that end, deprecate support for the three Intel CMT events in
>     "Rocky" release, with the intention to remove support for it in
>     the upcoming "Stein" release.  Note that we cannot deprecate /
>     remove `enabled_perf_events` config attribute altogether --
>     since there are other[+] `perf` events besides Intel CMT.
>     Whether anyone is using those other events with Nova is a good
>     question to which we don't have an equally good answer for, if
>     at all.
>
> Now we're removing[2] support for CMT events altogether.
>
> Question
> --------
>
> What I'm wondering now is the answer to the last sentence in the above
> quoted commit: "Whether anyone is using those other events with Nova is
> a good question to which we don't have an equally good answer for, if at
> all".
>
> If we know that "no one" (as if we can tell for sure) is using them, we
> can get rid of more dead code.
>
> So, any operators using the non-CMT events from here[0]?
>
> [0] https://libvirt.org/formatdomain.html#elementsPerf
> [1] https://opendev.org/openstack/nova/commit/fc4794acc6 â??libvirt:
>     Deprecate support for monitoring Intel CMT `perf` events
> [2] https://review.opendev.org/669129 â?? libvirt: Remove support for
>     Intel CMT `perf` event
>
> --
> /kashyap
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-discuss/attachments/20190719/865d5011/attachment.html>