| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-03-27 | |||
| 11:19:57 | sean-k-mooney | thats new to me | |
| 11:20:23 | sean-k-mooney | kashyap: is that syntax documented in oslo somewhere? | |
| 11:20:45 | kashyap | sean-k-mooney: Not quote, I found it from stephenfin's use here: https://github.com/openstack/nova/blob/cd15c3d/nova/conf/vnc.py#L226,L232 | |
| 11:20:50 | kashyap | s/quote/quite/ | |
| 11:20:59 | kashyap | sean-k-mooney: I only skimmed the oslo_config/cfg.py, though | |
| 11:21:21 | kashyap | https://github.com/openstack/oslo.config/blob/master/oslo_config/cfg.py | |
| 11:21:22 | sean-k-mooney | kashyap: well i have used choice elements for string fields before just never lists | |
| 11:21:44 | kashyap | sean-k-mooney: It's actually documented here: https://github.com/openstack/oslo.config/blob/master/oslo_config/cfg.py#L52 | |
| 11:24:09 | sean-k-mooney | ya so its not that you are using a listOpt that has a choice filed, you are instead create a list of sting enums which have the chioce element | |
| 11:24:14 | sean-k-mooney | but it works | |
| 11:24:55 | kashyap | Exactly :P It's still using the 'String' type | |
| 11:26:32 | openstackgerrit | Elod Illes proposed openstack/nova stable/ocata: WIP: Functional test for regression bug #1713783 https://review.openstack.org/505160 | |
| 11:26:33 | openstack | bug 1713783 in OpenStack Compute (nova) ocata "After failed evacuation the recovered source compute tries to delete the instance" [High,In progress] https://launchpad.net/bugs/1713783 - Assigned to Balazs Gibizer (balazs-gibizer) | |
| 11:32:14 | kashyap | sean-k-mooney: Thanks for the review. Much appreciated | |
| 11:33:10 | sean-k-mooney | kashyap: no worries. i dont get enough time as i like to review but always feel free to ping me with a patch if you have one | |
| 11:33:49 | kashyap | sean-k-mooney: Understood. I don't spend 100% time here either. I think I mostly know what topics pique your interest, will do :-) | |
| 11:48:26 | openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Noauth should also use request_id from compute_req_id.py https://review.openstack.org/555266 | |
| 12:09:27 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Allow scheduling only to enabled cells (Filter Scheduler) https://review.openstack.org/550527 | |
| 12:11:00 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Add --enable and --disable options to nova-manage update_cell https://review.openstack.org/555416 | |
| 12:12:39 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Update the cells FAQs and scheduler maintenance docs. https://review.openstack.org/556459 | |
| 12:13:17 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Update the cells FAQs and scheduler maintenance docs. https://review.openstack.org/556459 | |
| 12:35:29 | bhagyashris | Disallow rotation parameter 0 for 'createBackup' API) Request you to review the same. | |
| 12:35:29 | bhagyashris | johnthetubaguy: Hi, I have proposed revised (as per the discussion in Dublin PTG) spec: https://review.openstack.org/#/c/511825/2 ( | |
| 12:54:22 | efried | alex_xu_: Responded. | |
| 12:54:45 | bauzas | sorry folks, was at sports | |
| 12:57:37 | bauzas | sahid: thanks for having reviewed https://review.openstack.org/#/c/552924/ ! | |
| 12:58:49 | bauzas | sahid: you mean that page memories are different from the main memory ? | |
| 12:58:58 | efried | stephenfin: Just noticed you seem to have split personality according to stackalytics. It looks like your reviews go to one ID, your commits to the other. | |
| 12:59:12 | stephenfin | efried: Oh, really? | |
| 12:59:24 | efried | http://stackalytics.com/report/contribution/nova/90 -- sort by name and find thyself. | |
| 12:59:31 | bauzas | sahid: sorry, I'm not an expect for that, I just looked at https://docs.openstack.org/nova/pike/admin/huge-pages.html | |
| 12:59:47 | sahid | bauzas: i imagine when you are talking about main memroy, you are talking about small pages | |
| 13:00:12 | sahid | so if you allocate some huge pages, the small pages avail are going to decrease | |
| 13:00:34 | bauzas | sahid: okay, thanks | |
| 13:00:51 | bauzas | sahid: so, main memory and pages memories are different ? | |
| 13:01:03 | stephenfin | efried: Ah, one is using my email address, the other my launchpad ID. Guess I need to sync those somehow | |
| 13:01:05 | stephenfin | Or not. Meh | |
| 13:01:06 | bauzas | I thought it was just a size for eahc | |
| 13:01:58 | bauzas | stephenfin: efried: meh, when I see stackalytics, I'm sad :( | |
| 13:02:03 | sahid | bauzas: not sure what you mean but on a system that use hugepages that does not look right to tallk about main memory | |
| 13:02:10 | bauzas | because I don't have a lot of time for reviewing :( | |
| 13:02:15 | efried | I was like, wait, stephenfin has one commit in the last 90 days?? I *know* that ain't right. | |
| 13:02:21 | sahid | there are small pages and huge pages | |
| 13:02:28 | openstackgerrit | Theodoros Tsioutsias proposed openstack/nova-specs master: Enable rebuild for instances in cell0 https://review.openstack.org/554218 | |
| 13:02:53 | bauzas | sahid: okay, just to be clear, say I'm asking for a 1G page, it's different from asking to use 1GB for the main memory? | |
| 13:02:58 | openstackgerrit | Artom Lifshitz proposed openstack/nova-specs master: NUMA-aware live migration https://review.openstack.org/552722 | |
| 13:03:00 | bauzas | a different resource ? | |
| 13:03:03 | sahid | bauzas: yes | |
| 13:03:11 | bauzas | haha, thanks! | |
| 13:03:21 | bauzas | okay, I'll modify that then | |
| 13:05:35 | artom | bauzas only ^^^ for me | |
| 13:06:56 | Spaz-Home | Good luck on your specs day, my friends. | |
| 13:08:45 | stephenfin | sahid: RE: your concerns on the NUMA-aware vSwitches, I totally agree. It's a horrible hack | |
| 13:09:55 | stephenfin | sahid: However, it's the best we can get right now. Far as I can tell, there's no deterministic way to get this information from every vSwitch (at the moment, at least) and adding that would take some neutron plugins into compute-driver territory | |
| 13:10:59 | stephenfin | NUMA affinity for the OVS case is mostly driven from the PCI devices used, as from what I can tell OVS doesn't expose a "give me the NUMA affinity for this PCI device I've attached to my bridge" API. We'd need to implement this ourselves | |
| 13:11:17 | stephenfin | Which is something we really don't want to be doing in an ML2 driver, IMO | |
| 13:11:54 | stephenfin | The better model, which is what jaypipes, gibi and I discussed, was to use placement for this and collaboratively build up this model between nova and neutron. However, placement isn't there yet | |
| 13:12:28 | stephenfin | sahid: So this is making the best of a bad situation. If I've missed something though, definitely let me know. | |
| 13:12:35 | stephenfin | I'll put all the above in the review | |
| 13:16:13 | openstackgerrit | Sylvain Bauza proposed openstack/nova-specs master: Proposes NUMA topology with RPs https://review.openstack.org/552924 | |
| 13:21:38 | sahid | stephenfin: we don'-t need to ask that question "give me the NUMA affinity for this PCI device I've attached to my bridge" to OVS | |
| 13:22:04 | sahid | the operator is going to configure OVS and actually DPDK based on where the device is located | |
| 13:22:45 | sahid | so basically what we need is just to retrieve where the vhu ports are created | |
| 13:22:55 | sahid | this can be done by asking OVS | |
| 13:23:16 | sahid | so the neutron agent can return such information to nova by the binding detail od the port | |
| 13:23:20 | stephenfin | So just query the PMD pinning information for a bridge? | |
| 13:23:24 | sahid | do i have mised something? | |
| 13:24:38 | sahid | hum... no sure i underatand the PMD pinning for a bridge? | |
| 13:25:07 | stephenfin | It seemed like a big assumption to make (that PMD threads would be affined with the NIC), especially given that they don't have to be with recent releases | |
| 13:25:56 | sahid | well if operators want best perofrmance they have to do that | |
| 13:26:09 | stephenfin | Sorry, not the bridge. I'm referring to this https://developers.redhat.com/blog/2017/06/28/ovs-dpdk-parameters-dealing-with-multi-numa/ | |
| 13:26:14 | sahid | it's not our responsability (i don't think so) | |
| 13:26:27 | stephenfin | e.g. 'ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=0xF0' | |
| 13:26:45 | sahid | so that is configurd by operator | |
| 13:27:08 | sahid | we are expecting that the pmd to run on the same NUMA node where the phy NIC is, right? | |
| 13:27:41 | stephenfin | Yes. There can be multiple NICs | |
| 13:27:45 | stephenfin | ...too | |
| 13:28:11 | stephenfin | So NIC X is on NUMA node 0, while NIC Y is on NUMA node 1 | |
| 13:28:35 | sahid | but NIC X and NIC Y have different network | |
| 13:28:37 | sahid | right? | |
| 13:28:51 | stephenfin | and NIC X is connected to/tagged with physnet_x, and NIC Y to physnet_y | |
| 13:28:52 | stephenfin | right | |
| 13:29:10 | sahid | so when neutron is asking to create a port for X | |
| 13:29:38 | sahid | the agent can query OVS to know where that vhu is located? | |
| 13:29:54 | stephenfin | Can they? | |
| 13:30:02 | sahid | yep | |
| 13:30:08 | sahid | let a sec to find the command | |
| 13:30:27 | stephenfin | Right, I didn't know that :D | |
| 13:31:06 | alex_xu_ | efried: I missed one thing. If the resource provider X is the compute node. RP X provides VCPU and memory also. Each instance will consume the VCPU and memory. In that case, RP X will be returned | |
| 13:31:14 | sahid | stephenfin: https://software.intel.com/en-us/articles/vhost-user-numa-awareness-in-open-vswitch-with-dpdk | |
| 13:31:28 | sahid | if we can know here a port is located | |
| 13:31:31 | stephenfin | mikal: Just the man I'm looking for. Fancy jotting down your thoughts on https://review.openstack.org/554195 | |
| 13:31:38 | sahid | the agent can return this information to nova, no? | |
| 13:32:19 | stephenfin | (random aside: the OVS (-DPDK) documentation is hands-down awful. It sucks that we have to resort to random blogs for this critical information) | |
| 13:32:38 | stephenfin | sahid: If it's what we want then I don't see why not. Lemme check | |
| 13:33:42 | stephenfin | ooh, so 'pmd-rxq-show' does seem to be exactly what I wanted | |
| 13:35:06 | stephenfin | ...and we'd just assume that OVS was configured correctly so that there are PMD threads on all NUMA nodes, which is Red Hat's guidance at least | |
| 13:35:18 | stephenfin | sean-k-mooney[m]: If you're about, any thoughts on ^ | |
| 13:36:17 | efried | alex_xu_: Ah, then that's a mistake in modeling. | |
| 13:37:20 | efried | alex_xu_: The RP tree should not be set up such that the compute host provides inventory of FPGA. | |