Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-06-07
14:13:11 sdague my minimal config is
14:13:11 bauzas but yours is Neutron
14:13:17 sdague RABBIT_PASSWORD=pass
14:13:17 sdague DATABASE_PASSWORD=pass
14:13:17 sdague ADMIN_PASSWORD=pass
14:13:17 sdague [[local|localrc]]
14:13:19 sdague SERVICE_PASSWORD=pass
14:13:38 sdague all those networking things should not be required for working neutron anymore
14:14:10 kashyap sdague: Can you run this quick command -- `ps_mem` -- from inside the DevStack VM for me, and post the output somewhere?
14:14:22 dansmith bauzas: done
14:14:30 sdague kashyap: I'm not on a vm, I've got a real machine
14:14:37 sdague and ps_mem doesn't seem to be a command here
14:14:40 bauzas dansmith: cools
14:14:43 kashyap sdague: Sure, real machine does it, too
14:14:47 kashyap Wait let me get a URL for it
14:14:55 kashyap sdague: http://www.pixelbeat.org/scripts/ps_mem.py
14:15:06 dansmith bauzas: that one line of code has had like five bug fixes
14:15:12 dansmith bauzas: it's hall of fame stuff
14:15:22 kashyap sdague: The output would look something like - https://kashyapc.fedorapeople.org/virt/openstack/DevStack_Memory_Profile_29NOV2016_on_F25.txt
14:15:29 bauzas dansmith: which line ? :)
14:15:50 bauzas on the default_conn() ? heh
14:15:55 sdague http://paste.openstack.org/show/611722/
14:16:02 dansmith bauzas: just the calculation of the cell0 name I mean
14:16:09 bauzas dansmith: honestly, it's errorprone
14:16:29 bauzas given we need to support all the crazypants that operators can provide as strings
14:17:05 bauzas fun fact is that I verified, and I can actually use UTF-8 encoded creds with our current :)
14:17:19 bauzas if I was just providing the string like this
14:17:45 bauzas 'herv\xc3\xa9'
14:21:31 openstackgerrit Gábor Antal proposed openstack/nova master: Transform instance.live_migration_rollback notification https://review.openstack.org/402124
14:21:51 d3n14l Can someone tell me why a `openstack server list`repeatedly takes >15s for "just" 200 servers? (On openstack-ansible deployment, configured memcache and no slow request on the galera cluster…) See https://pastebin.com/nMivFhMA
14:23:55 openstackgerrit Gábor Antal proposed openstack/nova master: Transform instance.live_migration_rollback notification https://review.openstack.org/402124
14:23:59 kashyap sdague: Thank you
14:24:45 openstackgerrit Merged openstack/python-novaclient master: Prevent 2.32 release note from showing up in 9.0.0 release notes https://review.openstack.org/471383
14:24:49 kashyap sdague: Is this a "pure DevStack" host?
14:25:14 kashyap sdague: Because your setup is consuming 4.5 GB of memory; if you saw mine above, it was operating with 2.5GB
14:25:25 sdague kashyap: it is
14:25:34 sdague it's also been running for some amount of time
14:26:15 kashyap Noted; thanks
14:26:39 sdague while it never had a tempest run, I was doing some testing with guests booting so it's had about 30 - 40 cirros guests boot / delete
14:31:17 kashyap Ah
14:31:26 kashyap sdague: Any improvement with this: `free -m && sync && echo 3 > /proc/sys/vm/drop_caches && free -m`
14:31:32 kashyap As root
14:34:07 sdague kashyap: well that dumped my buffer cache
14:34:15 sdague but it doesn't change the ps_mem output
14:34:36 kashyap sdague: Whoops, thought you've read the line. That's why I didn't explain it, sorry :-(
14:34:49 kashyap But thank you for testing
14:35:05 sdague kashyap: what did you think I read?
14:35:25 kashyap The command sequence in quotes
14:35:25 sdague kashyap: what was your intent with the free calls?
14:35:48 sdague kashyap: sure, I guess I'm just not sure what you were asking for as "improvement"
14:35:51 kashyap Just to see if dropping 'needless' things in the caches will show what's the bare-minimal memory occupation
14:36:12 sdague but thats FS cache right?
14:36:27 sdague that's not going to impact any program mem usage
14:37:59 kashyap sdague: That's freeing page cache
14:38:10 sdague kashyap: yes
14:38:43 sdague kashyap: sorry, I guess I'm just not sure what you are asking when you asked about "improvement"
14:38:50 sdague what were you expecting?
14:38:56 sdague that you wanted me to verify
14:39:24 kashyap sdague: I think I erroneously thought it'll have effect on program memory usage
14:39:41 kashyap So just talking out loud, your brain spelled my stupidity out.
14:39:41 openstackgerrit Merged openstack/python-novaclient master: Updated from global requirements https://review.openstack.org/469581
14:47:13 openstackgerrit Jianghua Wang proposed openstack/nova master: Fix the race condition with novnc https://review.openstack.org/470798
15:00:35 openstackgerrit Andrey Volkov proposed openstack/nova master: [placement] Add api-ref for aggregates https://review.openstack.org/470932
15:00:35 openstackgerrit Andrey Volkov proposed openstack/nova master: [WIP] [placement] Add api-ref for allocations https://review.openstack.org/470933
15:08:08 openstackgerrit Baodong (Robert) Li proposed openstack/nova-specs master: Expose vlan trunking in metadata/configdrive https://review.openstack.org/471815
15:12:18 mriedem good lord there are just way too many config options for libvirt live migratoin
15:12:19 mriedem *migration
15:12:32 mriedem the help for some of them even reference things that no longer exist, which is helpful
15:14:39 openstackgerrit Gábor Antal proposed openstack/nova master: Move rebuild notification tests into separate method https://review.openstack.org/451789
15:20:00 kashyap mriedem: How many are you counting?
15:20:13 mriedem "a bazillion" is what my email to the ML is going to say
15:20:32 kashyap Heh, everything looks useful when you read one-by-one, doesn't it
15:21:00 mriedem sent
15:21:19 mriedem i took some artistic freedoms in language toward the end
15:21:24 kashyap :-)
15:21:54 kashyap But yeah, more seriously, exposing every frelling knob that libvirt has (which explicitly *doesn't* make policy decisions, by design) is not desirable
15:22:13 kashyap Must probaby turn some of them into defaults
15:23:30 mriedem s/landmine/minefield/ d'oh
15:23:45 mriedem well i'll see if i can cleanup some of these options
15:23:47 mriedem help-wise
15:24:02 dansmith landfill?
15:24:15 mriedem tire fire
15:24:54 jaypipes mriedem: there's way too many Nova config options for libvirt live migration because there's way too many configuration options for libvirt live migration. period...
15:25:52 mriedem the heart wants what it wants jay
15:25:58 mriedem and it wants mfing config options
15:26:10 kashyap The Nova time-outs and re-tries seemed to be the ones that can be removed?
15:26:25 gbarros mriedem I see that you +2 mhenkel's vif patch. Could you take a peek at the blueprint also pls?
15:26:40 gbarros mriedem https://blueprints.launchpad.net/nova/+spec/opencontrail-nova-vif-plugin
15:26:42 kashyap Err, before someone jumps on me, "removed" as in ... "can they be reconsiderd"
15:27:47 mriedem gbarros: as in approve the bp?
15:27:51 mriedem btw, -2 on the stable/ocata backport
15:27:56 mriedem we don't backport features
15:28:20 gbarros mriedem, on the bp, if it looks ok to you.
15:28:38 gbarros mriedem, ack on -2 for the backport. It was worth a shot :)
15:29:27 openstackgerrit Baodong (Robert) Li proposed openstack/nova-specs master: Expose vlan trunking in metadata/configdrive https://review.openstack.org/471815
15:29:57 gbarros mriedem, on the backport I'm mostly wanting to see if it will pass CI.
15:30:06 jaypipes kashyap: if there was a stable and consistently-converging live-migration interface in libvirt, those options wouldn't be needed. :)
15:30:09 mriedem gbarros: approved the bp, i forgot about that one
15:30:14 gbarros mriedem ty!
15:30:20 cdent jaypipes: i'm a bit confused and concerned about your lack of concern on the conusmer uuid thing, especially if the fix is doing and update or insert instead of what's there now?

Earlier   Later