From: Thomas Lamprecht <t.lamprecht@proxmox.com>
To: Proxmox VE development discussion <pve-devel@lists.proxmox.com>,
Fiona Ebner <f.ebner@proxmox.com>
Subject: Re: [pve-devel] [PATCH ha-manager 09/11] manager: use static resource scheduler when configured
Date: Wed, 16 Nov 2022 08:14:51 +0100 [thread overview]
Message-ID: <ff380260-8de5-57d8-321e-7a1e0b8893cf@proxmox.com> (raw)
In-Reply-To: <5aa83262-a8f4-8f8e-0612-55fbf9cfd7d9@proxmox.com>
Am 11/11/2022 um 10:28 schrieb Fiona Ebner:
> Am 10.11.22 um 15:37 schrieb Fiona Ebner:
>> @@ -206,11 +207,30 @@ my $valid_service_states = {
>> sub recompute_online_node_usage {
> So I was a bit worried that recompute_online_node_usage() would become
> too inefficient with the new add_service_usage_to_node() overhead from
> needing to read the guest configs. I now tested it with ~300 HA services
> (minimal containers) running on my virtual test cluster.
>
> Timings with 'basic' mode were between 0.0004 - 0.001 seconds
> Timings with 'static' mode were between 0.007 - 0.012 seconds
>
> While about a 10-fold increase, it's not too dramatic at least. I guess
> that's what the caching of cfs files is for :)
>
> Still, the function is currently not only called in the main loop in
> manage(), but also in next_state_recovery() and change_service_state().
>
> With, say, 400 HA services each on 5 nodes, if a node fails there's
> 400 calls from changing to freeze
huh, freeze should only happen on graceful shutdown of a node, not
if it fails?
> 400 calls from changing to recovery
> 400 calls in next_state_recovery
> 400 calls from changing to started
> If we take a generous estimate that each call takes 0.1 seconds (there's
> 2000 services in total), that's 40+80+40 seconds in 3 bursts during the
> fencing and recovery period.
doesn't that lead to overly long run windows between watchdog updates?
>
> Is that acceptable? Should I try to optimize how often the function is
> called?
>
hmm, a quick look wouldn't hurt, but not required for now IMO - if it can
interfere with watchdog updates I'd sneak in updating it once in between
though.
ps. maybe you can have some of that info/stats here in the commit message
of this patch.
next prev parent reply other threads:[~2022-11-16 7:15 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-10 14:37 [pve-devel] [PATCH-SERIES proxmox-resource-scheduling/pve-ha-manager/etc] add static usage scheduler for HA manager Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH proxmox-resource-scheduling 1/3] initial commit Fiona Ebner
2022-11-15 10:15 ` [pve-devel] applied: " Wolfgang Bumiller
2022-11-15 15:39 ` [pve-devel] " DERUMIER, Alexandre
2022-11-16 9:09 ` Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH proxmox-resource-scheduling 2/3] add pve_static module Fiona Ebner
2022-11-16 9:18 ` Thomas Lamprecht
2022-11-10 14:37 ` [pve-devel] [PATCH proxmox-resource-scheduling 3/3] add Debian packaging Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH proxmox-perl-rs 1/2] pve-rs: add resource scheduling module Fiona Ebner
2022-11-15 10:16 ` [pve-devel] applied-series: " Wolfgang Bumiller
2022-11-10 14:37 ` [pve-devel] [PATCH proxmox-perl-rs 2/2] add basic test for resource scheduling Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH manager 1/3] pvestatd: broadcast static node information Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH v3 manager 2/3] cluster resources: add cgroup-mode to node properties Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH v2 manager 3/3] ui: lxc/qemu: cpu edit: make cpuunits depend on node's cgroup version Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH cluster 1/1] datacenter config: add cluster resource scheduling (crs) options Fiona Ebner
2022-11-17 11:52 ` [pve-devel] applied: " Thomas Lamprecht
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 01/11] env: add get_static_node_stats() method Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 02/11] resources: add get_static_stats() method Fiona Ebner
2022-11-15 13:28 ` Thomas Lamprecht
2022-11-16 8:46 ` Fiona Ebner
2022-11-16 8:59 ` Thomas Lamprecht
2022-11-16 12:38 ` DERUMIER, Alexandre
2022-11-16 12:52 ` Thomas Lamprecht
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 03/11] add Usage base plugin and Usage::Basic plugin Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 04/11] manager: select service node: add $sid to parameters Fiona Ebner
2022-11-16 7:17 ` Thomas Lamprecht
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 05/11] manager: online node usage: switch to Usage::Basic plugin Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 06/11] usage: add Usage::Static plugin Fiona Ebner
2022-11-15 15:55 ` DERUMIER, Alexandre
2022-11-16 9:10 ` Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 07/11] env: add get_crs_settings() method Fiona Ebner
2022-11-16 7:05 ` Thomas Lamprecht
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 08/11] manager: set resource scheduler mode upon init Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 09/11] manager: use static resource scheduler when configured Fiona Ebner
2022-11-11 9:28 ` Fiona Ebner
2022-11-16 7:14 ` Thomas Lamprecht [this message]
2022-11-16 9:37 ` Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 10/11] manager: avoid scoring nodes if maintenance fallback node is valid Fiona Ebner
2022-11-10 14:37 ` [pve-devel] [PATCH ha-manager 11/11] manager: avoid scoring nodes when not trying next and current " Fiona Ebner
2022-11-10 14:38 ` [pve-devel] [PATCH docs 1/1] ha: add section about scheduler modes Fiona Ebner
2022-11-15 13:12 ` [pve-devel] partially-applied: [PATCH-SERIES proxmox-resource-scheduling/pve-ha-manager/etc] add static usage scheduler for HA manager Thomas Lamprecht
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ff380260-8de5-57d8-321e-7a1e0b8893cf@proxmox.com \
--to=t.lamprecht@proxmox.com \
--cc=f.ebner@proxmox.com \
--cc=pve-devel@lists.proxmox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox