Tuesday, 2017-08-22

*** thorst has quit IRC00:02
*** thorst has joined #openstack-powervm00:25
*** thorst has quit IRC00:26
*** esberglu has quit IRC00:33
*** openstackgerrit has joined #openstack-powervm00:42
openstackgerritOpenStack Proposal Bot proposed openstack/nova-powervm master: Updated from global requirements  https://review.openstack.org/49486000:42
*** thorst has joined #openstack-powervm01:26
*** thorst has quit IRC01:31
*** thorst has joined #openstack-powervm01:51
*** thorst has quit IRC01:51
*** thorst has joined #openstack-powervm02:52
*** thorst has quit IRC02:57
*** chhavi has joined #openstack-powervm03:52
*** thorst has joined #openstack-powervm03:53
*** thorst has quit IRC03:59
*** esberglu has joined #openstack-powervm04:21
*** esberglu has quit IRC04:31
*** thorst has joined #openstack-powervm04:54
*** thorst has quit IRC04:58
*** edmondsw has joined #openstack-powervm05:42
*** edmondsw has quit IRC05:47
*** thorst has joined #openstack-powervm05:55
*** thorst has quit IRC05:59
*** thorst has joined #openstack-powervm06:56
*** thorst has quit IRC07:00
*** edmondsw has joined #openstack-powervm07:30
*** edmondsw has quit IRC07:35
-openstackstatus- NOTICE: Gerrit is going to be restarted due to slow performance07:36
-openstackstatus- NOTICE: Gerrit has been restarted successfully07:41
*** k0da has joined #openstack-powervm07:54
*** thorst has joined #openstack-powervm07:57
*** thorst has quit IRC08:01
*** k0da has quit IRC08:37
*** k0da has joined #openstack-powervm08:50
*** thorst has joined #openstack-powervm08:57
*** thorst has quit IRC09:02
*** thorst has joined #openstack-powervm09:58
*** chhavi_ has joined #openstack-powervm10:02
*** thorst has quit IRC10:03
*** chhavi has quit IRC10:05
*** chhavi has joined #openstack-powervm10:06
*** chhavi_ has quit IRC10:06
*** AndyWojo has quit IRC11:01
*** zerick has quit IRC11:01
*** zerick has joined #openstack-powervm11:02
*** AndyWojo has joined #openstack-powervm11:03
*** chhavi_ has joined #openstack-powervm11:04
*** edmondsw has joined #openstack-powervm11:06
*** thorst has joined #openstack-powervm11:07
*** chhavi has quit IRC11:07
*** chhavi has joined #openstack-powervm11:08
*** chhavi_ has quit IRC11:10
*** edmondsw has quit IRC11:10
*** thorst has quit IRC11:11
*** svenkat has joined #openstack-powervm11:42
*** edmondsw has joined #openstack-powervm11:45
*** svenkat has quit IRC11:47
*** svenkat_ has joined #openstack-powervm11:47
*** k0da has quit IRC11:56
*** thorst has joined #openstack-powervm12:11
*** k0da has joined #openstack-powervm12:18
openstackgerritMerged openstack/nova-powervm master: Updated from global requirements  https://review.openstack.org/49486012:40
*** apearson has joined #openstack-powervm12:40
openstackgerritSridhar Venkat proposed openstack/nova-powervm master: WIP: Use OVS Attributes within pypowervm  https://review.openstack.org/48670212:44
thorstFYI - I'm going to have to miss the meeting today12:49
*** thorst is now known as thorst_afk12:49
*** kylek3h has joined #openstack-powervm12:50
*** esberglu has joined #openstack-powervm12:53
*** k0da has quit IRC12:58
esberglu#startmeeting powervm_driver_meeting13:00
openstackMeeting started Tue Aug 22 13:00:36 2017 UTC and is due to finish in 60 minutes.  The chair is esberglu. Information about MeetBot at http://wiki.debian.org/MeetBot.13:00
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.13:00
openstackThe meeting name has been set to 'powervm_driver_meeting'13:00
efriedjust getting through the morning rebases 8@13:01
esberglu#topic In Tree Driver13:01
esbergluI was planning on reviving the config drive WIP patch today13:02
esberglu#action esberglu: Update WIP config drive patch13:02
esbergluNot much else going on there atm right?13:03
esberglu#topic Out of Tree Driver13:03
esbergluAnything to report?13:04
mdrabeI'll have reviews for PPT ratio support out sometime this week after I live test13:04
mdrabeit for me13:04
efriedWhat's PPT ratio?13:05
mdrabepartition page table ratio13:05
efriedYeah, you're gonna have to spell that out at least once in the commit message, if not explain wtf it is.13:06
mdrabeLook at base_partition.py in that review for an explanation13:06
efriedCool.  Put that shit in the commit message.13:07
edmondswwe had another meeting about iSCSI13:07
edmondswsounds like we may be able to get some pvm changes going for iSCSI13:08
edmondswhave another meeting about that at the end of the week13:08
*** jay1_ has joined #openstack-powervm13:09
esbergluOkay cool. Move on to PCI then?13:10
efriedI opened a bug13:10
openstackLaunchpad bug 1711410 in nova-powervm "get_available_resource/get_inventory should account for resources used by the NovaLink, VIOSes, and non-OS partitions" [Undecided,New]13:10
efriedThat'll probably be mine, eventually, unless someone else wants to have a crack at it.13:11
efriedThat's all I have for OOT.13:13
esberglu#topic PCI Passthrough13:13
efriedProgress is being made.13:13
efriedI have been making chicken-scratch notes on what I've been doing.  Yesterday I put them into an etherpad: https://etherpad.openstack.org/p/powervm-pci-passthrough-notes13:14
efriedIt's mostly notes about how the nova code works.  At the bottom is some detail about changes I'm making to make stuff work.13:14
efriedOn my test system I've gotten to a point where 1) I can get a PCI claim to succeed and appear in the instance object to the spawn request; and 2) I can get it recognizing more than one type of PCI device at the same time.13:15
efriedTurns out the latter requires explicit unique PCI addresses, which have to be spoofed because PAL (Power Ain't Linux).13:15
efriedchangh has prototyped the REST code to return assigned partition UUID and type (and others) within the ManagedSystem IOSlot list.13:16
efriedNow that we've seen that PoC, I may rework 5749 to assume it's in place.13:17
openstackgerritSridhar Venkat proposed openstack/nova-powervm master: WIP: Use OVS Attributes within pypowervm  https://review.openstack.org/48670213:17
efriedThough it's likely to be next week before the REST function actually drops.13:17
efriedSo I'm going to wait on that and keep experimenting with sandbox code on my victim system.13:18
efriedbtw, for reference, the PCI address spoofing is 575513:18
efriedAlso for reference, the draft spec is here: https://review.openstack.org/49473313:18
efriedAnd the blueprint meta-doc is here: https://blueprints.launchpad.net/nova-powervm/+spec/pci-passthrough13:19
esbergluefried: Nice sounds like good progress! I still owe you another read through the spec13:19
efriedCurrently I'm working on a) Making sure I can claim multiple devices (homogenous and heterogeneous); and b) Doing the actual attach in spawn.13:20
efriedOn the table for discussion here...13:21
efriedDo we want to work this effort in nova proper instead of (or in parallel with) OOT13:22
efriedCon: it'll slow us way tf down.13:22
efriedPro: Exposure, and therefore increased likelihood of getting traction if we run up against something for which we require nova changes.13:23
edmondswefried I think we probably do want to work this in the communities vision13:25
*** k0da has joined #openstack-powervm13:25
edmondswwe can always do more/faster in parallel in nova-powervm, but if/when we need a community change it's going to be harder/slower to get that if they're out of the loop13:26
efriedOkay, so what I'll do is get a little further with the implementation, we can polish up the spec as a team, and then we can cut over and file a bp with nova.13:26
edmondswas for reworking your change to assume we have the stuff changh is adding, don't we need to also work with an older pvm level?13:27
edmondswok then... that certainly helps13:27
efriedWe will need to lockstep pypowervm releases, though, which is kind of a pita.13:28
efriedOkay, I think that's about all for PCI for now.  Still lots of work to do and a lot of unknowns.  But momentum is good.13:28
esberglu#topic PowerVM CI13:29
esbergluThings are looking a lot better with the timeout increase13:29
edmondswtx efried13:29
esbergluEspecially if you discount the runs that are failing due to that adapter delete/create serialization issue13:30
esbergluThe compute driver is still occasionally failing to start on some runs13:30
esbergluBut the n-cpu log is being cut short on those runs13:30
efriedalmost like the process is being killed?13:31
esbergluefried: Yeah either that or an issue with how we output the journald logs.13:31
esbergluBut I'm leaning towards the 1st13:32
esbergluBecause I've only seen the logs get cut off on the runs when the driver fails to come up properly13:32
efriedand we fixed the journald business by stopping the services13:32
efriedthough that actually shouldn't be necessary at all.13:33
efriedand probably had no effect :)13:33
esbergluYou can see all of the other services get stopped but not n-cpu13:33
esbergluI'll investigate that further this pm and see if I can get some more information13:34
edmondswis anything happening on the serialization issue?13:35
esbergluThere's a bug open for it. I can ask again today13:35
esbergluThat's all I had for CI13:36
efriedchangh has been mentioning in scrums that he's been looking into it.13:36
efriedI believe we've all agreed there is a bug at or below REST at this point.13:36
edmondswgood... I thought they were pushing back and wasn't sure how seriously they were taking it13:37
efriedWe should keep pestering changh during scrums, make sure it stays in the forefront.13:37
esberglu#topic Driver Testing13:38
esberglujay1_: Anything new on your end?13:38
jay1_esberglu: wanted to understand ISCSI issue fixing ?13:41
jay1_edmondsw:Also for S3 I could see there are two stories present13:42
edmondswjay1_ you're supposed to be testing vscsi and fc this sprint13:42
edmondswwe're working on getting iSCSI fixed, but that will be a process involving pvm rest changes13:43
jay1_ah okay..13:43
edmondswjay1_ fyi, I've added several things to the stacking issues etherpad13:44
edmondswmake sure you don't stumble over any of those13:44
edmondswjay1_ https://etherpad.openstack.org/p/powervm_stacking_issues13:45
jay1_edmondsw can you add more info to the stories like in the acceptance13:45
edmondswjay1_ yes13:46
jay1_OOT + VSCSI , you mean with SSP ?13:46
jay1_is it like we have to run Tempest scripts or manual effort is also required ?13:47
edmondswjay1_ no... vSCSI is an alternative to SSP13:47
edmondswjay1_ I'd prefer tempest13:47
edmondswso you'd setup an environment using vSCSI instead of SSP and then run tempest13:48
jay1_okay.. how about pike and queens13:49
thorst_afkvSCSI is a connection type.  I think vSCSI for SSP for the nova boot volume.  vSCSI for FC for the cinder volume13:49
thorst_afkor just boot from Cinder (though that's sometimes harder)13:49
thorst_afkjust my 2 cents13:49
esberglu#topic Open Discussion13:51
esbergluAnything else today?13:51
edmondswjay1_ the stories for this sprint are pike13:51
jay1_edmondsw sure13:52
jay1_thorst_afk last release NPIV with FC we wanted to do, same even now ?13:53
edmondswjay1_ that's the 2nd story13:54
edmondswone for vSCSI, one for NPIV13:54
openstackMeeting ended Tue Aug 22 13:58:58 2017 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)13:59
openstackMinutes:        http://eavesdrop.openstack.org/meetings/powervm_driver_meeting/2017/powervm_driver_meeting.2017-08-22-13.00.html13:59
openstackMinutes (text): http://eavesdrop.openstack.org/meetings/powervm_driver_meeting/2017/powervm_driver_meeting.2017-08-22-13.00.txt13:59
openstackLog:            http://eavesdrop.openstack.org/meetings/powervm_driver_meeting/2017/powervm_driver_meeting.2017-08-22-13.00.log.html13:59
thorst_afkjay1_: yes, as edmondsw noted14:06
jay1_Ok.. so with VSCSI we have to try both SSP novaboot and FC cinder boot ?14:06
edmondswthorst_afk I'm still trying to puzzle out your vSCSI connection type statement...14:07
edmondswit sounds like you meant SSP boot vol, vSCSI data vol, but I don't know if that works... at least PowerVC doesn't support it14:07
openstackgerritSridhar Venkat proposed openstack/nova-powervm master: Use OVS Attributes within pypowervm  https://review.openstack.org/48670214:07
thorst_afkedmondsw: vSCSI is just how you connect a disk to a VM.  PVC land screws up the term.  SSP uses vSCSI to connect the 'SSP disk to the VM'.  vSCSI can also connect a PV (fibre channel physical volume) to the VM14:07
edmondswright... stil trying to figure out what you meant :)14:08
thorst_afkI'm just saying that vSCSI doesn't mean Cinder.14:09
thorst_afkvSCSI is used for everything (except NPIV cinder disks)14:09
edmondswso... SSP boot, cinder (vSCSI) data?14:10
edmondswI don't think that works14:10
thorst_afkit does14:10
edmondswok... PowerVC definitely doesn't support it14:10
edmondswjay1_ ok, did you follow that?14:11
edmondswyou can use SSP for boot volumes, but use cinder for data volumes14:11
edmondswfor one of the two stories, use vSCSI FC for the data volumes, for the other use NPIV FC for the data volumes14:12
edmondswI'll put this in the acceptance tab14:12
jay1_so, boot is common from SSP right ?14:13
thorst_afkjay1_: yeah, I think that's fine.  That's *easiest*14:13
thorst_afkbecause you put the image into glance, then it pushes that image to SSP14:14
thorst_afkif you boot from volume, you'd need to seed the image in the volume...and it'd be a pain14:14
thorst_afkyou just need to make sure your glance image supports RMC, so that you can attach volumes14:14
jay1_Yeah.. will use our Ubuntu image which we used in the previous release.14:15
*** dwayne has joined #openstack-powervm15:08
*** kylek3h has quit IRC15:20
*** kylek3h has joined #openstack-powervm15:39
*** dwayne has quit IRC15:43
*** thorst_afk has quit IRC15:55
*** thorst_afk has joined #openstack-powervm15:57
*** thorst_afk has quit IRC16:01
*** thorst_afk has joined #openstack-powervm16:05
*** dwayne has joined #openstack-powervm16:09
*** miltonm has quit IRC16:28
*** apearson has quit IRC16:39
*** apearson has joined #openstack-powervm16:39
*** thorst_afk has quit IRC17:13
*** apearson has quit IRC17:24
*** apearson has joined #openstack-powervm17:25
*** thorst_afk has joined #openstack-powervm17:49
*** thorst_a_ has joined #openstack-powervm18:18
*** thorst_afk has quit IRC18:20
* efried just got PCI claims for multiple devices (2 of one kind and 1 of another)18:45
* efried is ready to start assigning devices18:46
*** miltonm has joined #openstack-powervm18:56
*** jay1_ has quit IRC19:01
*** apearson has quit IRC19:02
*** apearson has joined #openstack-powervm19:07
efriedmiltonm Are you an expert on DRC index perchance?19:40
*** apearson has quit IRC19:41
miltonmefried, not really, more platform design ... I know they exist, I know they come from the device tree which uses 32 bit, I know there are hypercalls that take indexes in 64 bit registers19:41
miltonmmore I track hardware and how the os interacts with it ...19:42
efriedmiltonm Sigh, okay.19:42
esbergluedmondsw: efried: thorst_a_: Do we need to cut releases for *-powervm for 1.3.3 FP1? Looks like there's 1 commit to ocata for nova-powervm and networking-powervm since the last ocata tag19:44
esbergluNone for ceilometer-powervm19:44
*** apearson has joined #openstack-powervm19:45
efriedesberglu I don't think so, but others may disagree.19:46
edmondswesberglu nobody's asked me about that. Send me links for those commits and I'll ask19:47
esbergluedmondsw: ^ links above19:47
* edmondsw wonders how he missed that...19:47
*** miltonm has quit IRC19:47
*** apearson has quit IRC19:52
*** apearson has joined #openstack-powervm19:53
edmondswesberglu I assume we would need releases, but checking... odd that they haven't asked19:54
efriedceph: https://review.openstack.org/#/c/484478/20:08
*** tjakobs has joined #openstack-powervm20:08
edmondswesberglu apparently PowerVC 1.3.3 patched the nova-powervm RPM they ship/install to include that because we were having CI issues and they couldn't wait20:11
edmondswI'm checking the networking-powervm one now20:12
efriedtjakobs Still can't move forward with the ceph code until pypowervm 1.1.7 is cut and through g-r.20:13
tjakobsefried any idea when that happens?20:13
edmondswesberglu yeah, sounds like the same on the networking-powervm one20:18
edmondswso unless thorst_a_ thinks we should go ahead and cut releases in case someone else needs these... because PowerVC already has them20:19
efriedtjakobs Whenever we decide, I suppose.  Any special hurry to get the ceph stuff done?20:36
tjakobsefried Not that I'm aware of.20:38
*** chhavi has quit IRC20:47
*** apearson has quit IRC21:15
*** k0da has quit IRC21:20
*** thorst_a_ has quit IRC21:21
*** thorst_afk has joined #openstack-powervm21:24
*** thorst_afk has quit IRC21:28
*** k0da has joined #openstack-powervm21:36
*** k0da has quit IRC21:50
*** svenkat_ has quit IRC21:55
*** dwayne has quit IRC22:01
*** thorst_afk has joined #openstack-powervm22:02
*** k0da has joined #openstack-powervm22:06
*** thorst_afk has quit IRC22:06
*** thorst_afk has joined #openstack-powervm22:08
*** edmondsw has quit IRC22:17
*** thorst_afk has quit IRC22:18
*** thorst_afk has joined #openstack-powervm22:18
*** thorst_afk has quit IRC22:22
*** tjakobs has quit IRC22:43
openstackgerritEric Fried proposed openstack/nova-powervm master: WIP: PCI passthrough  https://review.openstack.org/49643422:57
*** thorst_afk has joined #openstack-powervm23:00
*** dwayne has joined #openstack-powervm23:04
*** miltonm has joined #openstack-powervm23:18
*** esberglu has quit IRC23:24
*** esberglu has joined #openstack-powervm23:25
*** esberglu has quit IRC23:29
*** efried has quit IRC23:42
*** k0da has quit IRC23:42
*** edmondsw has joined #openstack-powervm23:55
*** efried has joined #openstack-powervm23:56

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!