<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Apr 6, 2018 at 2:54 PM, Chris Dent <span dir="ltr"><<a href="mailto:cdent+os@anticdent.org" target="_blank">cdent+os@anticdent.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
This is "contract" style update. New stuff will not be added to the<br>
lists.<br>
<br>
# Most Important<br>
<br>
There doesn't appear to be anything new with regard to most<br>
important. That which was important remains important. At the<br>
scheduler team meeting at the start of the week there was talk of<br>
working out ways to trim the amount of work in progress by using the<br>
nova priorities tracking etherpad to help sort things out:<br>
<br>
<a href="https://etherpad.openstack.org/p/rocky-nova-priorities-tracking" rel="noreferrer" target="_blank">https://etherpad.openstack.org<wbr>/p/rocky-nova-priorities-track<wbr>ing</a><br>
<br>
Update provider tree and nested allocation candidates remain<br>
critical basic functionality on which much else is based. With most<br>
of provider tree done, it's really on nested allocation candidates.<br>
<br>
# What's Changed<br>
<br>
Quite a bit of provider tree related code has merged.<br>
<br>
Some negotiation happened with regard to when/if the fixes for<br>
shared providers is going to happen. I'm not sure how that resolved,<br>
if someone can follow up with that, that would be most excellent.<br>
<br>
Most of the placement-req-filter series merged.<br>
<br>
The spec for error codes in the placement API merged (code is in<br>
progress and ready for review, see below).<br>
<br>
# Questions<br>
<br>
* Eric and I discussed earlier in the week that it might be a good<br>
time to start an #openstack-placement IRC channel, for two main<br>
reasons: break things up so as to limit the crosstalk in the often<br>
very busy #openstack-nova channel and to lend a bit of momentum<br>
for going in that direction. Is this okay with everyone? If not,<br>
please say so, otherwise I'll make it happen soon.<br>
<br></blockquote><div><br></div><div>Fine by me. It's sometimes difficult to follow all the conversations so having a separate channel looks good to me, at least for discussing only about specific Placement questions.<br></div><div>For Nova related points (like how to use nested RPs for example with NUMA), maybe #openstack-nova is still the main IRC channel for that.<br></div><div> <br><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
* Shared providers status?<br>
(I really think we need to make this go. It was one of the<br>
original value propositions of placement: being able to accurate<br>
manage shared disk.)<br>
<br>
# Bugs<br>
<br>
* Placement related bugs not yet in progress: <a href="https://goo.gl/TgiPXb" rel="noreferrer" target="_blank">https://goo.gl/TgiPXb</a><br>
15, -1 on last week<br>
* In progress placement bugs: <a href="https://goo.gl/vzGGDQ" rel="noreferrer" target="_blank">https://goo.gl/vzGGDQ</a><br>
13, +1 on last week<br>
<br>
# Specs<br>
<br>
These seem to be divided into three classes:<br>
<br>
* Normal stuff<br>
* Old stuff not getting attention or newer stuff that ought to be<br>
abandoned because of lack of support<br>
* Anything related to the client side of using nested providers<br>
effectively. This apparently needs a lot of thinking. If there are<br>
some general sticking points we can extract and resolve, that<br>
might help move the whole thing forward?<br>
<br>
* <a href="https://review.openstack.org/#/c/549067/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/549067/</a><br>
VMware: place instances on resource pool<br>
(using update_provider_tree)<br>
<br>
* <a href="https://review.openstack.org/#/c/545057/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/545057/</a><br>
mirror nova host aggregates to placement API<br>
<br>
* <a href="https://review.openstack.org/#/c/552924/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/552924/</a><br>
Proposes NUMA topology with RPs<br>
<br>
* <a href="https://review.openstack.org/#/c/544683/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/544683/</a><br>
Account for host agg allocation ratio in placement<br>
<br>
* <a href="https://review.openstack.org/#/c/552927/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/552927/</a><br>
Spec for isolating configuration of placement database<br>
(This has a strong +2 on it but needs one more.)<br>
<br>
* <a href="https://review.openstack.org/#/c/552105/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/552105/</a><br>
Support default allocation ratios<br>
<br>
* <a href="https://review.openstack.org/#/c/438640/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/438640/</a><br>
Spec on preemptible servers<br>
<br>
* <a href="https://review.openstack.org/#/c/556873/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/556873/</a><br>
Handle nested providers for allocation candidates<br>
<br>
* <a href="https://review.openstack.org/#/c/556971/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/556971/</a><br>
Add Generation to Consumers<br>
<br>
* <a href="https://review.openstack.org/#/c/557065/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/557065/</a><br>
Proposes Multiple GPU types<br>
<br>
* <a href="https://review.openstack.org/#/c/555081/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/555081/</a><br>
Standardize CPU resource tracking<br>
<br>
* <a href="https://review.openstack.org/#/c/502306/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/502306/</a><br>
Network bandwidth resource provider<br>
<br>
* <a href="https://review.openstack.org/#/c/509042/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/509042/</a><br>
Propose counting quota usage from placement<br>
<br>
# Main Themes<br>
<br>
## Update Provider Tree<br>
<br>
Most of the main guts of this have merged (huzzah!). What's left are<br>
some loose end details, and clean handling of aggregates:<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/update-provider-tree" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/update-provider-tr<wbr>ee</a><br>
<br>
## Nested providers in allocation candidates<br>
<br>
Representing nested provides in the response to GET<br>
/allocation_candidates is required to actually make use of all the<br>
topology that update provider tree will report. That work is in<br>
progress at:<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/nested-resource-providers" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/nested-resource-pr<wbr>oviders</a><br>
<a href="https://review.openstack.org/#/q/topic:bp/nested-resource-providers-allocation-candidates" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/nested-resource-pr<wbr>oviders-allocation-candidates</a><br>
<br>
Note that some of this includes the up-for-debate shared handling.<br>
<br>
## Request Filters<br>
<br>
As far as I can tell this is mostly done (yay!) but there is a loose<br>
end: We merged an updated spec to support multiple member_of<br>
parameters, but it's not clear anybody is currently owning that:<br>
<br>
<a href="https://review.openstack.org/#/c/555413/" rel="noreferrer" target="_blank">https://review.openstack.org/<wbr>#/c/555413/</a><br>
<br>
## Mirror nova host aggregates to placement<br>
<br>
This makes it so some kinds of aggregate filtering can be done<br>
"placement side" by mirroring nova host aggregates into placement<br>
aggregates.<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/placement-mirror-host-aggregates" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/placement-mirror-h<wbr>ost-aggregates</a><br>
<br>
It's part of what will make the req filters above useful.<br>
<br>
## Forbidden Traits<br>
<br>
A way of expressing "I'd like resources that do _not_ have trait X".<br>
This is ready for review:<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/placement-forbidden-traits" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/placement-forbidde<wbr>n-traits</a><br>
<br>
## Consumer Generations<br>
<br>
This allows multiple agents to "safely" update allocations for a<br>
single consumer. There is both a spec and code in progress for this:<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/add-consumer-generation" rel="noreferrer" target="_blank">https://review.openstack.org/<wbr>#/q/topic:bp/add-consumer-gene<wbr>ration</a><br>
<br>
# Extraction<br>
<br>
Small bits of work on extraction continue on the<br>
bp/placement-extract topic:<br>
<br>
<a href="https://review.openstack.org/#/q/topic:bp/placement-extract" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/placement-extract</a><br>
<br>
The spec for optional database handling got some nice support<br>
but needs more attention:<br>
<br>
<a href="https://review.openstack.org/#/c/552927/" rel="noreferrer" target="_blank">https://review.openstack.org/<wbr>#/c/552927/</a><br>
<br>
Jay has declared that he's going to start work on the<br>
os-resources-classes library.<br>
<br>
I've posted a 6th in my placement container playground series:<br>
<br>
<a href="https://anticdent.org/placement-container-playground-6.html" rel="noreferrer" target="_blank">https://anticdent.org/placemen<wbr>t-container-playground-6.html</a><br>
<br>
Though not directly related to extraction, that experimentation has<br>
exposed a lot of the areas where work remains to be done to make<br>
placement independent of nova.<br>
<br>
A recent experiment with shrinking the repo to just the placement<br>
dir reinforced a few things we already know:<br>
<br>
* The placement tests need their own base test to avoid 'from nova<br>
import test'<br>
* That will need to provide database and other fixtures (such a<br>
config and the self.flags feature).<br>
* And, of course, eventually, config handling. The container<br>
experiments above demonstrate just how little config placement<br>
actually needs (by design, let's keep it that way).<br>
<br>
# Other<br>
<br>
This is a contract week, so nothing new has been added here, despite<br>
there being new work. Part of the intent here it make sure we are<br>
queue-like where we can be. This list maintains its ordering from<br>
week to week: newly discovered things are added to the end.<br>
<br>
There are 14 entries here, -7 on last week.<br>
<br>
That's good. However some of the removals are the result of some<br>
code changing topic (and having been listed here by topic). Some of<br>
the oldest stuff at the top of the list has not moved.<br>
<br>
* <a href="https://review.openstack.org/#/c/546660/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/546660/</a><br>
Purge comp_node and res_prvdr records during deletion of<br>
cells/hosts<br>
<br>
* <a href="https://review.openstack.org/#/q/topic:bp/placement-osc-plugin-rocky" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/q/topic:bp/placement-osc-plug<wbr>in-rocky</a><br>
A huge pile of improvements to osc-placement<br>
<br>
* <a href="https://review.openstack.org/#/c/546713/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/546713/</a><br>
Add compute capabilities traits (to os-traits)<br>
<br>
* <a href="https://review.openstack.org/#/c/524425/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/524425/</a><br>
General policy sample file for placement<br>
<br>
* <a href="https://review.openstack.org/#/c/546177/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/546177/</a><br>
Provide framework for setting placement error codes<br>
<br>
* <a href="https://review.openstack.org/#/c/527791/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/527791/</a><br>
Get resource provider by uuid or name (osc-placement)<br>
<br>
* <a href="https://review.openstack.org/#/c/477478/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/477478/</a><br>
placement: Make API history doc more consistent<br>
<br>
* <a href="https://review.openstack.org/#/c/556669/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/556669/</a><br>
Handle agg generation conflict in report client<br>
<br>
* <a href="https://review.openstack.org/#/c/556628/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/556628/</a><br>
Slugification utilities for placement names<br>
<br>
* <a href="https://review.openstack.org/#/c/557086/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/557086/</a><br>
Remove usage of [placement]os_region_name<br>
<br>
* <a href="https://review.openstack.org/#/c/556633/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/556633/</a><br>
Get rid of 406 paths in report client<br>
<br>
* <a href="https://review.openstack.org/#/c/537614/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/537614/</a><br>
Add unit test for non-placement resize<br>
<br>
* <a href="https://review.openstack.org/#/c/554357/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/554357/</a><br>
Address issues raised in adding member_of to GET /a-c<br>
<br>
* <a href="https://review.openstack.org/#/c/493865/" rel="noreferrer" target="_blank">https://review.openstack.org/#<wbr>/c/493865/</a><br>
cover migration cases with functional tests<br>
<br>
# End<br>
<br>
2 runway slots open up this coming Wednesday, the 11th.<span class="HOEnZb"><font color="#888888"><br>
<br>
-- <br>
Chris Dent ٩◔̯◔۶ <a href="https://anticdent.org/" rel="noreferrer" target="_blank">https://anticdent.org/</a><br>
freenode: cdent tw: @anticdent</font></span><br>______________________________<wbr>______________________________<wbr>______________<br>
OpenStack Development Mailing List (not for usage questions)<br>
Unsubscribe: <a href="http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe" rel="noreferrer" target="_blank">OpenStack-dev-request@lists.<wbr>openstack.org?subject:<wbr>unsubscribe</a><br>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" rel="noreferrer" target="_blank">http://lists.openstack.org/<wbr>cgi-bin/mailman/listinfo/<wbr>openstack-dev</a><br>
<br></blockquote></div><br></div></div>