Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-11-16
14:26:49 johnthetubaguy mriedem: can I help?
14:27:29 jaypipes tssurya: there is no result from a DELETE statement. unless you're referring to the number of records deleted?
14:28:07 jaypipes mriedem: DiskFilter only looks at local_gb_used/local_gb AFAIK
14:28:13 openstackgerrit Claudiu Belu proposed openstack/nova master: doc: Adds Hyper-V PCI passthrough details https://review.openstack.org/510467
14:28:16 jaypipes mriedem: which obviously isn't placement...
14:28:25 openstackgerrit Hamdy Khader proposed openstack/nova master: Adding NVMEoF for libvirt driver https://review.openstack.org/482640
14:28:36 tssurya jaypipes : yes well we know that there is only the number of records that we get, but we were wondering if there was a way to get the records too ?
14:30:38 jaypipes tssurya: not unless you do a SELECT first, then DELETE.
14:30:49 jaypipes tssurya: in the same read-consistent transaction of course.
14:31:08 tssurya jaypipes : we have some integrity issues with the list of records deleted and inserted during archiving, so I was looking for an efficient way to do this.
14:31:17 tssurya jaypipes : ok! thanks :)
14:31:33 jaypipes tssurya: np :)
14:33:43 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (usage) https://review.openstack.org/520603
14:40:58 openstackgerrit Hamdy Khader proposed openstack/nova master: Adding NVMEoF for libvirt driver https://review.openstack.org/482640
14:43:03 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (trait) https://review.openstack.org/520605
14:47:44 openstackgerrit Balazs Gibizer proposed openstack/nova master: Test allocation candidates: multiple aggregates https://review.openstack.org/518633
14:47:44 openstackgerrit Balazs Gibizer proposed openstack/nova master: Test alloc_cands with one RP shared between two RPs https://review.openstack.org/519617
14:47:45 openstackgerrit Balazs Gibizer proposed openstack/nova master: Test helper: validate provider summaries https://review.openstack.org/518982
14:49:00 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (aggregate) https://review.openstack.org/520608
14:55:32 mriedem johnthetubaguy: jaypipes: my assertion was that the DiskFilter won't work with shared storage on the computes
14:55:39 mriedem so shouldn't be enabled if you're using NFS, Ceph, etc
14:56:00 johnthetubaguy its kinda useless in that case, yeah
14:56:12 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (resource_class) https://review.openstack.org/520611
15:04:39 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (inventory) https://review.openstack.org/520613
15:08:28 mriedem pretty simple novaclient change to add support for microversion 2.54 https://review.openstack.org/#/c/519572/
15:17:57 openstackgerrit Ed Leafe proposed openstack/nova master: Refactor the code to check for sufficient hosts https://review.openstack.org/520242
15:17:57 openstackgerrit Ed Leafe proposed openstack/nova master: Add Selection objects https://review.openstack.org/499239
15:17:58 openstackgerrit Ed Leafe proposed openstack/nova master: Modify select_destinations() to return objects and alts https://review.openstack.org/510159
15:17:58 openstackgerrit Ed Leafe proposed openstack/nova master: Return Selection objects from the scheduler driver https://review.openstack.org/495854
15:17:59 openstackgerrit Ed Leafe proposed openstack/nova master: Move the claim_resources method to scheduler utils https://review.openstack.org/511357
15:17:59 openstackgerrit Ed Leafe proposed openstack/nova master: Change RPC for select_destinations() https://review.openstack.org/516707
15:18:00 openstackgerrit Ed Leafe proposed openstack/nova master: Make conductor pass and use host_lists https://review.openstack.org/511358
15:21:38 mriedem yikun: https://review.openstack.org/#/c/515640/
15:22:39 jianghuaw_ johnthetubaguy, hi John. Could you help to look at this XenAPI fix? https://review.openstack.org/#/c/514902/
15:24:07 jianghuaw_ johnthetubaguy, I think that's a straightforward fix.
15:29:46 mriedem edleafe: are you going to make the changes in https://review.openstack.org/#/c/520242/ and fix the tests on the change after it? i'd like to hit those today while i'm still on the trail
15:29:52 mriedem the alternate hosts trail
15:29:53 mriedem of blood
15:33:50 edleafe mriedem: you mean the >= change?
15:34:17 edleafe if we ever select more hosts than requested, we have much bigger problems
15:35:11 mriedem ok so leave that,
15:35:15 mriedem but there are the typos and unrelated change
15:35:18 mriedem easy fixes
15:35:33 mriedem i figured you'd hit those while you cleanup the test failures in the next patch above
15:36:23 edleafe mriedem: you did see that I pushed an update to the series like 20 minutes ago?
15:37:15 mriedem orly
15:37:17 mriedem i didn't :)
15:37:51 edleafe :)
15:39:01 mriedem ok, one little thing in the test
15:42:04 johnthetubaguy jianghuaw_: been meaning to look at that, its on my list
15:43:07 openstackgerrit Ed Leafe proposed openstack/nova master: Refactor the code to check for sufficient hosts https://review.openstack.org/520242
15:43:15 edleafe mriedem: ^ added
15:44:42 mriedem thanks, lgtm
15:44:51 mriedem jaypipes: bauzas: you should review https://review.openstack.org/#/c/520242/ as well
15:44:58 edleafe mriedem: the series is blocked because you (plural) wanted to verify the whole process before beginning to commit any of the series. Is it possible to get such a review so that we can begin to merge some of these lower patches?
15:45:43 mriedem jaypipes: bauzas: review note, this change is because the next one in the series is going to be returning Selection objects from FilterScheduler._schedule to FilterScheduler.select_destinations, where that HostState.updated = None was done, which won't work with a Selection object
15:45:57 mriedem so ed moved that validation down into the _schedule part before we convert HostState to Selection objects
15:46:03 jaypipes understood.
15:46:22 mriedem edleafe: i'm working through the series
15:47:17 edleafe mriedem: I understand. I'm thinking of a more high-level view of the end result to verify the approach. Each piece can be later reviewed for specific issues
15:47:36 edleafe I'd just like to merge the lower patches to keep rebase issues to a minimum
15:47:48 edleafe Especially since I'll be out all next week
15:48:12 mriedem i can't promise anything
15:48:29 mriedem but it's good to know you'll be out next week so i don't ping you incessantly about this
15:50:03 mriedem that reminds me, need to request wednesday off...
15:51:17 cfriesen mdbooth: you around?
15:51:56 mdbooth cfriesen: I am
15:53:02 jianghuaw_ johnthetubaguy, thanks:-)
15:54:10 cfriesen mdbooth: we ran into an issue with "qemu-img convert" for a large image on a compute node consuming a lot of cache and affecting the performance of instances on that compute node. we fixed it by adding "-t none" to have it bypass the cache.
15:54:31 openstackgerrit Matt Riedemann proposed openstack/python-novaclient master: [ci] Use pseudo-random names for new resources https://review.openstack.org/479777
15:54:46 cfriesen mdbooth: do you think this is something that would be useful more generally?
15:54:51 kashyap cfriesen: "-t" == cache?
15:55:03 cfriesen kashyap: yes.
15:55:18 kashyap cfriesen: And you had 'writeback'?
15:55:18 mdbooth Hmm
15:55:42 cfriesen kashyap: no, normally nova doesn't specify anything for -t so you just get the default
15:55:54 kashyap The default is writeback :-)
15:56:10 kashyap It's the QEMU default that libvirt / Nova inherit
15:56:12 cfriesen well, okay, but we didn't explicitly specify it.
15:57:19 cfriesen _disk_qcow2_to_raw() in nova/virt/libvirt/driver.py
15:57:30 cfriesen and vice-versa
15:57:57 mdbooth kashyap: Yes, cfriesen is essentially complaining that writeback isn't what he wants.
15:58:44 mdbooth cfriesen: What impact did this have on the machine?
15:59:26 kashyap mdboth: I see. In that case, I know cfriesen sometimes posts on 'qemu-devel'; so Chris if you have the perf data then post there and Cc 'qemu-block' list...
15:59:31 efried jaypipes Got local changes to the n-r-p series? I'm doing a restack, and can hit Tetsuro's comments on https://review.openstack.org/#/c/384807/ while I'm at it.
15:59:48 kashyap But yeah, I'll let him to describe the issue here
15:59:49 jaypipes efried: nope.
15:59:52 mdbooth I'm wondering if we give instances cache=none by default
15:59:55 efried jaypipes ight
15:59:59 cfriesen kashyap: mdbooth: digging up the details
16:00:01 mdbooth I thought we did?
16:00:14 kashyap mdbooth: Easy to check, let's examine a Gate job logs...
16:00:41 kashyap I thought we let the default be 'writeback' as QEMU gives us; because there's a reason they settled on it as the default
16:02:10 mdbooth kashyap: Are you sure writeback is the default?
16:02:13 mdbooth writeback is terrible.
16:02:17 kashyap mdbooth: Yes, I just checked
16:02:31 kashyap mdbooth: Just fire up `man qemu-system-x86` and look for 'writeback'
16:03:01 kashyap mdbooth: And...Here's the evidence from the gate jobs:
16:03:04 kashyap http://logs.openstack.org/14/407514/9/check/legacy-grenade-dsvm-neutron-multinode-live-migration/8d355c4/logs/libvirt/qemu/instance-00000019.txt.gz
16:05:10 cfriesen mdbooth: kashyap: from our bug notes it looks like we were actually interfering with other host processes, not the guests. (my bad, I misremembered) So we turned off the cache and limited it to two disk-intensive operations (glance downloads, image conversions, etc.) at a time.

Earlier   Later