| *** _hemna has joined #openstack-cinder | 00:05 | |
| *** tkajinam has quit IRC | 00:10 | |
| *** lixiaoy1 has joined #openstack-cinder | 00:32 | |
| *** jdillaman has quit IRC | 00:37 | |
| *** jdillaman has joined #openstack-cinder | 00:38 | |
| *** spatel has joined #openstack-cinder | 00:46 | |
| *** Liang__ has joined #openstack-cinder | 00:50 | |
| *** _hemna has quit IRC | 00:55 | |
| *** eharney has quit IRC | 00:57 | |
| *** TxGirlGeek has joined #openstack-cinder | 00:58 | |
| *** enriquetaso has quit IRC | 01:17 | |
| *** deiter has quit IRC | 01:18 | |
| *** imacdonn has quit IRC | 01:18 | |
| *** imacdonn has joined #openstack-cinder | 01:18 | |
| *** TxGirlGeek has quit IRC | 01:32 | |
| *** whoami-rajat has joined #openstack-cinder | 01:43 | |
| *** baojg has joined #openstack-cinder | 01:45 | |
| *** _erlon_ has quit IRC | 01:46 | |
| *** TxGirlGeek has joined #openstack-cinder | 01:50 | |
| *** TxGirlGeek has quit IRC | 01:54 | |
| *** Kuirong has joined #openstack-cinder | 01:57 | |
| *** ruffian_sheep has joined #openstack-cinder | 02:09 | |
| ruffian_sheep | whoami-rajat:I am sorry that I am asking you every day. However, due to the many reasons involved, the integration of this drive is more important to our company. I need to report the situation of cinder entering the main line on a daily basis. | 02:09 |
|---|---|---|
| Kuirong | @ruffian_sheep I think they cannot reply immediately becaz live in different time zones. | 02:15 |
| whoami-rajat | ruffian_sheep: hey, I got involved with some other task so couldn't fully review the patch. Apologies for that. Will surely provide feedback till EOD. | 02:17 |
| Kuirong | Deadline is 7/26 right? | 02:19 |
| whoami-rajat | Right | 02:22 |
| *** spatel has quit IRC | 02:34 | |
| ruffian_sheep | whoami-rajat:I am very grateful to receive your reply as soon as possible. | 02:42 |
| ruffian_sheep | Kuirong :I also asked him to go online before asking him.lol | 02:43 |
| ruffian_sheep | Does it mean that 7.26 stops all the main lines of this time? | 02:47 |
| Kuirong | I think so. | 02:50 |
| ruffian_sheep | Kuirong :Is the main line integration drive semi-annual? | 02:53 |
| Kuirong | yes | 03:00 |
| ruffian_sheep | Kuirong :Thx | 03:06 |
| *** psachin has joined #openstack-cinder | 03:27 | |
| *** gkadam has joined #openstack-cinder | 03:49 | |
| *** gkadam has quit IRC | 03:50 | |
| *** ruffian_sheep has quit IRC | 03:51 | |
| *** dviroel has quit IRC | 03:52 | |
| *** tejdeep has quit IRC | 03:55 | |
| *** udesale has joined #openstack-cinder | 04:01 | |
| *** rcernin has quit IRC | 04:13 | |
| *** rcernin has joined #openstack-cinder | 04:14 | |
| *** rcernin has quit IRC | 04:20 | |
| *** pcaruana has joined #openstack-cinder | 04:43 | |
| *** Luzi has joined #openstack-cinder | 05:05 | |
| *** m75abrams has joined #openstack-cinder | 05:11 | |
| *** TxGirlGeek has joined #openstack-cinder | 05:16 | |
| *** TxGirlGeek has quit IRC | 05:43 | |
| *** dpawlik has joined #openstack-cinder | 05:44 | |
| openstackgerrit | Brin Zhang proposed openstack/cinder master: Rollback the quota_usages table when failed to create a incremental backup https://review.opendev.org/626790 | 06:24 |
| *** tejdeep has joined #openstack-cinder | 06:38 | |
| *** bhagyashris has joined #openstack-cinder | 06:39 | |
| *** e0ne has joined #openstack-cinder | 06:40 | |
| *** e0ne has quit IRC | 06:41 | |
| *** georgeakahiron has joined #openstack-cinder | 06:47 | |
| *** sahid has joined #openstack-cinder | 07:02 | |
| *** tesseract has joined #openstack-cinder | 07:09 | |
| *** irclogbot_1 has quit IRC | 07:20 | |
| *** kaisers has quit IRC | 07:20 | |
| *** openstackstatus has quit IRC | 07:20 | |
| *** irclogbot_2 has joined #openstack-cinder | 07:21 | |
| *** kaisers has joined #openstack-cinder | 07:21 | |
| *** dansmith has quit IRC | 07:23 | |
| *** sahid has quit IRC | 07:23 | |
| *** dansmith has joined #openstack-cinder | 07:24 | |
| *** sahid has joined #openstack-cinder | 07:24 | |
| *** Anticimex has quit IRC | 07:24 | |
| *** tosky has joined #openstack-cinder | 07:28 | |
| *** Anticimex has joined #openstack-cinder | 07:29 | |
| *** tejdeep has quit IRC | 07:42 | |
| *** tejdeep_ has joined #openstack-cinder | 07:42 | |
| *** tejdeep_ has quit IRC | 08:03 | |
| *** sahid has quit IRC | 08:06 | |
| *** sahid has joined #openstack-cinder | 08:09 | |
| *** lixiaoy1 has quit IRC | 08:28 | |
| *** e0ne has joined #openstack-cinder | 08:46 | |
| openstackgerrit | Brin Zhang proposed openstack/cinder master: Rollback the quota_usages table when failed to create a incremental backup https://review.opendev.org/626790 | 08:47 |
| *** sapd1_x has joined #openstack-cinder | 08:55 | |
| openstackgerrit | Merged openstack/cinder master: Update api-ref location https://review.opendev.org/672087 | 09:10 |
| *** _hemna has joined #openstack-cinder | 09:12 | |
| *** _hemna has quit IRC | 09:16 | |
| *** Liang__ has quit IRC | 09:19 | |
| *** psachin has quit IRC | 09:20 | |
| *** ociuhandu has joined #openstack-cinder | 09:31 | |
| *** psachin has joined #openstack-cinder | 09:35 | |
| *** georgeakahiron has quit IRC | 09:40 | |
| *** jojoda has quit IRC | 09:45 | |
| *** ociuhandu has quit IRC | 09:48 | |
| openstackgerrit | Brin Zhang proposed openstack/cinder master: Rollback the quota_usages table when failed to create a incremental backup https://review.opendev.org/626790 | 09:48 |
| *** ociuhandu has joined #openstack-cinder | 09:50 | |
| *** bhagyashris has quit IRC | 09:54 | |
| *** dpawlik has quit IRC | 10:02 | |
| *** dpawlik has joined #openstack-cinder | 10:04 | |
| *** sahid has quit IRC | 10:10 | |
| *** lpetrut has joined #openstack-cinder | 10:21 | |
| *** brinzhang_ has joined #openstack-cinder | 10:26 | |
| *** brinzhang has quit IRC | 10:30 | |
| *** sahid has joined #openstack-cinder | 10:47 | |
| *** spatel has joined #openstack-cinder | 11:20 | |
| *** spatel has quit IRC | 11:25 | |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder master: QNAP: Fix login on Python3 https://review.opendev.org/672265 | 11:25 |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder master: QNAP: Avoid unnecessary sleeps https://review.opendev.org/672269 | 11:46 |
| *** carloss has joined #openstack-cinder | 11:47 | |
| *** markvoelker has quit IRC | 11:58 | |
| *** eharney has joined #openstack-cinder | 12:01 | |
| *** udesale has quit IRC | 12:04 | |
| *** udesale has joined #openstack-cinder | 12:04 | |
| *** raghavendrat has joined #openstack-cinder | 12:11 | |
| *** dviroel has joined #openstack-cinder | 12:14 | |
| *** markvoelker has joined #openstack-cinder | 12:16 | |
| *** henriqueof has joined #openstack-cinder | 12:33 | |
| *** mriedem has joined #openstack-cinder | 12:38 | |
| *** deiter has joined #openstack-cinder | 12:53 | |
| *** enriquetaso has joined #openstack-cinder | 12:57 | |
| *** raghavendrat has quit IRC | 12:59 | |
| hemna | mornin | 13:19 |
| *** openstackstatus has joined #openstack-cinder | 13:21 | |
| *** ChanServ sets mode: +v openstackstatus | 13:21 | |
| jungleboyj | Morning. | 13:22 |
| BLZbubba | hello | 13:22 |
| openstackgerrit | Pawel Kaminski proposed openstack/os-brick master: connectors/nvme: Wait utill nvme device show up in kernel https://review.opendev.org/672031 | 13:27 |
| *** davidsha has joined #openstack-cinder | 13:28 | |
| *** mriedem has quit IRC | 13:38 | |
| *** lemko has joined #openstack-cinder | 13:38 | |
| openstackgerrit | Pawel Kaminski proposed openstack/cinder master: target/spdknvmf: Add configuration parameter https://review.opendev.org/672064 | 13:39 |
| *** tosky_ has joined #openstack-cinder | 13:40 | |
| *** tosky has quit IRC | 13:42 | |
| *** Luzi has quit IRC | 13:51 | |
| geguileo | mnaser: I forgot to mention yesterday that for Active-Active you'll also have to configure TooZ in Cinder to use a DLM (like etcd with a gateway) | 13:54 |
| *** tosky_ is now known as tosky | 14:08 | |
| *** jrubenst has joined #openstack-cinder | 14:10 | |
| *** baojg has quit IRC | 14:18 | |
| *** baojg has joined #openstack-cinder | 14:19 | |
| *** mriedem has joined #openstack-cinder | 14:36 | |
| *** dosaboy has joined #openstack-cinder | 14:41 | |
| openstackgerrit | Bhaa Shakur proposed openstack/cinder master: Zadara VPSA: Move to API access key authentication https://review.opendev.org/670715 | 14:43 |
| *** _hemna has joined #openstack-cinder | 14:47 | |
| *** _hemna has quit IRC | 14:52 | |
| *** TxGirlGeek has joined #openstack-cinder | 14:52 | |
| *** lseki has joined #openstack-cinder | 15:06 | |
| *** enriquetaso has quit IRC | 15:09 | |
| *** enriquetaso has joined #openstack-cinder | 15:09 | |
| *** jc_ has joined #openstack-cinder | 15:11 | |
| jc_ | Hi, I have installed OpenStack with one controller and two compute nodes. I am trying to install block storage on both compute nodes (compute + block). Is it possible, or Do i need to separate them? | 15:13 |
| *** enriquetaso has quit IRC | 15:14 | |
| *** trident has quit IRC | 15:18 | |
| *** _hemna has joined #openstack-cinder | 15:18 | |
| *** vanou has joined #openstack-cinder | 15:19 | |
| hemna | heh huawei laid off 2/3 of US employees | 15:19 |
| hemna | poof | 15:19 |
| *** trident has joined #openstack-cinder | 15:20 | |
| jungleboyj | On top of all the ones that left? | 15:20 |
| hemna | yup | 15:20 |
| jungleboyj | Yikes. Not totally surprised I guess. | 15:21 |
| * jungleboyj will not get political | 15:21 | |
| jungleboyj | Glad I work for Lenovo. | 15:21 |
| jungleboyj | hemna: Any word on your Mid-Cycle attendance? | 15:22 |
| BLZbubba | sad | 15:22 |
| BLZbubba | jc_: do you mean you're trying to converge a cinder-lvm instance on each compute node? | 15:23 |
| smcginnis | Oh wow. | 15:23 |
| smcginnis | Hadn't heard that yet. | 15:23 |
| hemna | jungleboyj: I should be able to attend | 15:24 |
| jc_ | BLZbubba: yes. But I am not successful just following the cinder documentation. | 15:25 |
| BLZbubba | i don't see why it wouldn't work... | 15:26 |
| jungleboyj | hemna: Awesome. We will have at least have you, me, eharney and rosmaita there then. | 15:26 |
| BLZbubba | personally i would suggest making a vm for cinder and passing the storage & networking in via virtual functions... this should reduce the chance of weird nova/cinder conflicts | 15:27 |
| jc_ | as per the cinder install doc (https://docs.openstack.org/cinder/stein/install/cinder-storage-install-ubuntu.html), I have installed the cinder, but I am not able to see my complete storage in the local storage of admin panel | 15:28 |
| BLZbubba | i run ceph osd as a vm like this on my openstack nodes | 15:28 |
| BLZbubba | socket 0 for ceph, socket 1 for qemu (nova) | 15:28 |
| *** dpawlik has quit IRC | 15:30 | |
| *** m75abrams has quit IRC | 15:32 | |
| smcginnis | jc_: Does "cinder service-list" show your two cinder-volume services? | 15:32 |
| jc_ | smcginnis: yes , on controller node it shows that both the nodes are up. | 15:34 |
| smcginnis | jc_: Do you have a volume type for each backend and can you create a volume from each? | 15:34 |
| jc_ | smcginnis: you mean, I need to create volume on each server. I tried with vg and lvextended with all the physical disks, but i did not see complete local storage in horizon | 15:37 |
| *** e0ne has quit IRC | 15:39 | |
| smcginnis | jc_: Not sure what you mean now. You should have had a volume group on each host that you configured cinder to use when installing the service there. You shouldn't need to do any other lvm commands after that initial setup of the space you want to use. | 15:40 |
| *** david-lyle is now known as dklyle | 15:40 | |
| jc_ | smcginnis: https://paste.ubuntu.com/p/C2cNmmxQB6/, I have added the details from two servers on paste bin. | 15:42 |
| *** sapd1_x has quit IRC | 15:44 | |
| jc_ | I have created volume group on each host. But my problem is, I am not able to see the added storage in the neutron. When I am creating the VMs, I THINK the created VMs are not using the cinder storage space. | 15:45 |
| vanou | jungleboyj: hello. I want to confirm your comment on 3rd party ci at last weekly mtg. | 15:46 |
| vanou | you said | 15:46 |
| vanou | will be going through the CI results in the next couple of weeks to see who all is running with Py3.7. I have seen at least one py3.5. I am not going to mark a driver unsupported for that, but by the end of Train it will need to be py3.7 or it will get marked. | 15:46 |
| vanou | Dose this mean 'Failing to make ci running py3.7 till milestone 2 dosen't mark a driver unsupported, but failing to pass py37 test & to running py3.7 in ci env till 12th Sep marks a driver unsupported' ? | 15:47 |
| vanou | Forgive my long posting, this is first time of IRC. | 15:48 |
| smcginnis | jc_: You mean horizon, not neutron, right? | 15:49 |
| smcginnis | Are you trying to boot a VM from a Cinder volume or add a volume to a VM? | 15:49 |
| jungleboyj | vanou: Hey, let me read here. | 15:50 |
| smcginnis | vanou: Yes, all third party CI will need to be running services under py3.7. If not done by that later date, it will be marked unsupported. | 15:50 |
| vanou | jungleboyj: thanks! | 15:51 |
| jungleboyj | vanou: So, if you are trying to run py37 but it isn't passing yet. A patch will not be pushed up to mark you unsupported. | 15:51 |
| jungleboyj | It will be marked unsupported at milestone 3 if py37 isn't passing by then. | 15:51 |
| vanou | smcginnis: Thanks your comment! Does 'that later date' mean '12th Sep'? | 15:51 |
| *** trident has quit IRC | 15:52 | |
| jungleboyj | If py3 isn't running at all, then it will be marked unsupported after this week. | 15:52 |
| *** _hemna has quit IRC | 15:53 | |
| jc_ | smcginnis: sorry, I mean in Horizon. I want the VMs to use the cinder volume, when they are created. As , I am not seeing the total storage in the horizon, I think the volume added to cinder-volumes is not available for VM creation or there is something I am missing or don't know at the moment. | 15:53 |
| smcginnis | jc_: So you have one cinder node configured to use h018-vg in cinder.conf and the other to use and cinder-volumes in the other cinder.conf? | 15:53 |
| vanou | jungleboy: Sorry. What is difference between "if you are trying to run py37 but it isn't passing yet." and "If py3 isn't running at all" ? | 15:55 |
| *** trident has joined #openstack-cinder | 15:55 | |
| smcginnis | jc_: Then you created two volume types, one for each backend? In horizon you can either create the volume first from the Volumes section, or you can create the volume when you create the VM. The drawback with the latter is you can't specify which volume type to use and you need to have a default volume type set so nova can create the cinder volume. | 15:56 |
| jc_ | smcginnis: I can change the cinder volume names to same on both hosts, but How can i fix the storage available for usage. | 16:00 |
| vanou | jungleboyj: Sorry. I misspelled your nick. What is difference between "if you are trying to run py37 but it isn't passing yet." and "If py3 isn't running at all" ? Because I understand 'this week' in "If py3 isn't running at all, then it will be marked unsupported after this week." as "this week (Jul 21 - 27 2019)". | 16:00 |
| smcginnis | jc_: What volume names are you changing? You shouldn't be modifying any volumes created via cinder. | 16:01 |
| *** e0ne has joined #openstack-cinder | 16:01 | |
| jc_ | smcginnis: these names (h018-vg and cinder-volumes) are created by me as discussed in https://docs.openstack.org/cinder/stein/install/cinder-storage-install-ubuntu.html. They were not created by cinder. | 16:06 |
| jungleboyj | vanou: If you have a job producing logs showing you are trying to test py37, you are good. If you aren't running any tests running py37, then it will be unsupported. | 16:06 |
| jungleboyj | So, if for patches you are running py27 and it is passing CI and you are doing a separate run with py37 but it is still failing, that is ok until Milestone3. | 16:07 |
| smcginnis | jc_: That name of the volume group doesn't matter as long as that's what you've configured in that node's cinder.conf to use. | 16:07 |
| jungleboyj | It is probably easiest to just switch to using py37 though. | 16:07 |
| *** enriquetaso has joined #openstack-cinder | 16:07 | |
| *** tesseract has quit IRC | 16:07 | |
| *** lemko has quit IRC | 16:08 | |
| openstackgerrit | Chris M proposed openstack/cinder master: Fix dothill multiattach support https://review.opendev.org/671942 | 16:08 |
| openstackgerrit | Chris M proposed openstack/cinder master: Create Seagate driver from dothill driver https://review.opendev.org/671195 | 16:08 |
| vanou | jungleboyj: I understand. Should I post logs to mailing list or so to show trying to run py37? | 16:10 |
| jc_ | smcginnis: ok, how can i allocate more space for VM creation ( have more volume to allocate when VM is created from horizon) | 16:10 |
| jungleboyj | Are you able to have the logs show up with your CI run? | 16:11 |
| jungleboyj | vanou: ^^ | 16:11 |
| smcginnis | It looks like you should have 50+ TB available on each node. Does it tell you there is no space on either one? | 16:11 |
| smcginnis | jc_: Oh wait, you've created your own volume in both of those volume groups using all the space it looks like according to that paste from earlier. | 16:12 |
| *** tejdeep has joined #openstack-cinder | 16:12 | |
| vanou | jungleboyj: Not yet... | 16:12 |
| vanou | But I'll... | 16:12 |
| jungleboyj | Which driver is this for? | 16:15 |
| jc_ | smcginnis: I have around 60TB + 60 TB on each host, but on the horizon I can only see 350 + 350 GB (where OS is installed for both hosts). Everything else is shown in the volume groups, but I am not able to add more storage for VM creation. | 16:15 |
| vanou | Fujitsu Eternus | 16:15 |
| geguileo | hemna: how is your customer setting the iSCSI boot? | 16:16 |
| hemna | ? | 16:16 |
| openstackgerrit | Luigi Toscano proposed openstack/cinder master: Port the legacy multibackend jobs to Zuul v3 https://review.opendev.org/671945 | 16:16 |
| geguileo | hemna: the customer with the problem of the iSCSI boot volume and the OS-Brick | 16:16 |
| pots | jungleboyj: smcginnis: if i have two patches submitted (multi-attach fix and seagate driver), and they conflict with each other, should i make one dependent on the other or just make them all relative to master and rebase them as needed until they are all merged? it seems like zuul won't apply the dependencies in order to validate the dependent pa | 16:16 |
| pots | tch? | 16:16 |
| smcginnis | jc_: You're using all the space with the volumes you've created from those volume groups. | 16:17 |
| *** sahid has quit IRC | 16:17 | |
| * hemna looks over teh bug | 16:17 | |
| smcginnis | pots: I would just rebase as needed. | 16:17 |
| jungleboyj | smcginnis: ++ | 16:18 |
| *** lpetrut has quit IRC | 16:19 | |
| *** enriquetaso has quit IRC | 16:19 | |
| jc_ | smcginnis: so I should not add all the drives in lvm, so that I can mount some of the hdd and use them for instances? | 16:21 |
| whoami-rajat | geguileo: Hi, could you take a look at the fix https://review.opendev.org/#/c/670887/ . Thanks! | 16:22 |
| smcginnis | jc_: No, that part is right. All drives (at least any that you want to use) should be added to the volume group | 16:22 |
| smcginnis | jc_: But the volumes you've created from the volume groups are using most of your space, so not sure why you are doing that. If that was a mistake, you should just delete those logical volumes. | 16:23 |
| *** baojg has quit IRC | 16:26 | |
| vanou | jungleboyj: Sorry. Fujitsu Eternus | 16:28 |
| jungleboyj | vanou: Ok. Cool. Thank you for keeping up to date on this. | 16:28 |
| jc_ | smcginnis: so, I should not use all the HDD for vgcreate rather, I should choose only few disks for cinder volume group. What should i do with the remaining HDD, should mounted them to be used for VM instances at /var/lib/nova/instances ? | 16:29 |
| smcginnis | jc_: No. All drives you want to use in cinder should be added to the vg. | 16:30 |
| vanou | jungleboyj: Thanks your kind & patience! It is midnight in my timezone. So, I'll leave. Thanks! | 16:31 |
| jungleboyj | Ok. Thanks for keeping us updated. | 16:31 |
| jungleboyj | Have a good night. | 16:31 |
| *** davidsha has quit IRC | 16:32 | |
| *** vanou has quit IRC | 16:34 | |
| jc_ | smcginnis: but the drives added to cinder, are not shown in Horizon. On different setup, I have mounted some of the HDD at /var/lib/nova/instances by adding them to a VG. In that scenario, I saw the local storage on Horizon. (This setup doesn't exist now) | 16:34 |
| smcginnis | jc_: Because you've created a logical volume from the volume group using up most of your space. | 16:35 |
| jc_ | smcginnis: sorry, but I have checked all my commands. I have not created logical volume on one of my host, but cinder created it seems. Other host (I have created) | 16:40 |
| *** deiter has quit IRC | 16:40 | |
| jc_ | smcginnis: I am not sure, how to fix this problem. | 16:40 |
| smcginnis | Cinder created sda on h018-vg? That doesn't seem right. I don't think cinder would set a name like that. | 16:41 |
| geguileo | whoami-rajat: reviewed | 16:43 |
| smcginnis | Wow, our LVM documentation appears to really be lacking any kind of useful detail. | 16:44 |
| jc_ | smcginnis: I created h018, that is old host. I have installed on a different host today where I just created pvcreate /dev/sdX ... ... and then vgcreate cinder-volumes /dev/sdX ... ... lsblk output from host where cinder created (maybe) https://paste.ubuntu.com/p/4wfn6Xd4DV/ | 16:45 |
| smcginnis | jc_: Ah, I see the one is a thin pool. So make sure the cinder.conf on that node is pointing to that pool, not the VG as I had previously said. | 16:50 |
| rosmaita | smcginnis: jungleboyj: eharney: geguileo: e0ne: jgriffith: and anyone else interested in stable branch maintenance policy | 16:50 |
| smcginnis | jc_: h018-vg/sda does not appear to be a thin pool though. | 16:50 |
| rosmaita | this is a follow up to last week's discussion about the stable/rocky release situation | 16:50 |
| rosmaita | we were deciding what to do about https://review.opendev.org/#/c/639867/ "Declare multiattach support for HPE MSA" | 16:50 |
| rosmaita | pots agreed to do some testing to see if multiattach does work for HPE MSA | 16:50 |
| rosmaita | turns out that there are some bugfixes needed | 16:50 |
| rosmaita | for example, https://review.opendev.org/#/c/671942/5 | 16:51 |
| rosmaita | so, I think declaring multiattach is a bit premature | 16:51 |
| rosmaita | my concern is that there are 21 unreleased changes in stable/queens that are being blocked | 16:51 |
| rosmaita | so my proposal is to restore https://review.opendev.org/#/c/670086/ to revert the declaration in stable/rocky | 16:51 |
| rosmaita | then release 13.0.6 for rocky and 12.0.8 for queens | 16:51 |
| rosmaita | after that, we can continue backporting the dothill fixes definitely to stein, and perhaps farther | 16:51 |
| rosmaita | (pots should probably hold off on the renaming to seagate until we have this stuff working and backported) | 16:51 |
| rosmaita | so let's discuss here, or i can write this up and send to the ML for discussion if people are busy now | 16:51 |
| smcginnis | rosmaita: I think we should probably just revert that patch then. If it was just a flag being returned that was fine for stable. If it required some additional code changes, then I don't think we should have done it. | 16:51 |
| whoami-rajat | geguileo: Thanks. I thought that was the base structure for implementing tests and db/memory test files were helper. | 16:52 |
| whoami-rajat | Also i think i tried setting the PERSISTENCE_CFG to memory but ran into failures, see PS3. | 16:52 |
| whoami-rajat | Will try again after the refactoring. | 16:52 |
| rosmaita | smcginnis: that's what i was thinking too | 16:52 |
| geguileo | whoami-rajat: it's the base class for testing persistence plugins | 16:52 |
| geguileo | whoami-rajat: but in your patch you want to test the helper methods | 16:52 |
| eharney | rosmaita: reverting sounds like a good idea to me | 16:53 |
| geguileo | whoami-rajat: aka, I did a poor job with the classes :-( | 16:53 |
| pots | i'd like to continue with the dothill->seagate rename though. is there any reason not to? | 16:53 |
| geguileo | whoami-rajat: and now you are paying for it, sorry | 16:53 |
| e0ne | smcginnis: +1 to revert it | 16:53 |
| rosmaita | pots: only reason would be if it complicates backports | 16:53 |
| rosmaita | i think we still need to have the discussion about driver backports at the midcycle | 16:54 |
| jungleboyj | I am ok with doing the revert given the additional info. | 16:54 |
| rosmaita | because i think it's important to backport bugfixes for these | 16:54 |
| *** enriquetaso has joined #openstack-cinder | 16:54 | |
| jungleboyj | rosmaita: ++ | 16:54 |
| whoami-rajat | geguileo: oh, no issues. Also it makes the addition of future tests easier. Thanks for the help. :) | 16:55 |
| jungleboyj | pots: Yeah, that is the concern with the driver rename. | 16:55 |
| geguileo | whoami-rajat: thank you for working on the fix | 16:55 |
| *** gnufied has joined #openstack-cinder | 16:55 | |
| rosmaita | we just need to formulate a coherent policy that the stable-maint team will accept | 16:55 |
| jungleboyj | If others agree, I think we could allow delaying the rename a bit to allow for you to do the fix backport more easily. | 16:55 |
| pots | i'm not sure what the difference is, though, whether we rename the driver now or in the future? | 16:55 |
| jc_ | smcginnis: i should change it in [DEFAULT] or [lvm] , it should changed to "volume_group = cinder-volumes-pool" , is that correct? | 16:56 |
| jungleboyj | Backporting the changes from master are more complicated once the name has changed. | 16:56 |
| *** tejdeep has quit IRC | 16:56 | |
| rosmaita | i think you will have a merge conflict on every backport to stein | 16:57 |
| smcginnis | jc_: The LVM pool is set under the backend section, so if you've named that [lvm], that's where. Not under [DEFAULT]. | 16:57 |
| rosmaita | ok, so as far as the stable/rocky release goes, it looks like we have a majority of cinder stable-maint cores saying revert the change | 16:59 |
| rosmaita | i will revise the patches | 16:59 |
| rosmaita | if anyone has second thoughts, you can leave comments on the patches | 16:59 |
| pots | ok, in that case should the seagate driver just be another subclass of the dothill driver? or should it just be a new standalone driver, which the other drivers can subclass in the future, at which point we can just remove the old dothill code. That way, there are no renames to complicate backports. | 17:01 |
| jungleboyj | pots: Well, if you subclass it, then you don't complicate things for backporting changes. | 17:01 |
| jungleboyj | That or you go forward with the rename and deal with the merge conflicts any time you need to backport a fix. | 17:02 |
| rosmaita | i would be against subclassing it | 17:03 |
| pots | yes, but won't there always be an argument that we shouldn't rneame the dothill code because of the merge conflicts with backports? if we introduce the new driver first, there won't be any merge conflicts with backports, just sometimes duplicate patches if we have to patch both drivers. | 17:03 |
| *** udesale has quit IRC | 17:03 | |
| jungleboyj | I think the argument is always going to be with the subclassing. | 17:04 |
| rosmaita | yeah, it's one of those things where if we have a bunch of bugfixes that need to be done right now, lets fix first & then rename | 17:04 |
| rosmaita | but if it's just this one patch, then probably not a big deal | 17:04 |
| jungleboyj | That was why I was proposing we let you delay the rename to get the patch in and then go from there. | 17:05 |
| pots | my concern is missing the deadline for adding the seagate driver. | 17:07 |
| rosmaita | gotcha | 17:07 |
| *** trident has quit IRC | 17:07 | |
| rosmaita | that is a legitimate concern | 17:08 |
| pots | i have a seagate-ci ready to start posting results with py37 but for some problems getting the multiattach tests to pass | 17:08 |
| jc_ | smcginnis: changed it and restarted nova-compute and cinder-volume. It still shows the same. | 17:08 |
| jungleboyj | pots: Right, that was why I was trying to see if the team would be ok with letting that slide a bit given the situation. | 17:08 |
| *** trident has joined #openstack-cinder | 17:10 | |
| jungleboyj | I can't make that decision alone though. :-) | 17:11 |
| pots | i'm fine with whatever works for the group, just need you to point me in the right direction. | 17:11 |
| rosmaita | jungleboyj: what are your thoughts exactly on "sliding"? | 17:12 |
| jungleboyj | I am saying that we can slip the requirement to get the Seagate change in by a week or two while he get this other issue resolved. | 17:13 |
| *** ociuhandu has quit IRC | 17:13 | |
| jungleboyj | Given that it isn't a NEW driver. It is a rebrand. | 17:13 |
| jungleboyj | We have allowed those to go in later in the past I think. | 17:13 |
| rosmaita | i personally think that is fine -- it doesn't set much of a precedent because it's a pretty unique case | 17:14 |
| rosmaita | i think as far as backporting fixes, reviews are easier if the backports are clean | 17:14 |
| rosmaita | so i think it would be a good tradeoff | 17:14 |
| jungleboyj | Okie dokie. | 17:16 |
| *** lseki has quit IRC | 17:16 | |
| jungleboyj | pots: You ok with that? | 17:19 |
| pots | sure. shall i submit a seagate driver as a subclass for now and we can hold off on renaming the underlying dothill driver until the next cycle? | 17:20 |
| *** tejdeep has joined #openstack-cinder | 17:20 | |
| jungleboyj | No. Lets just get the patches to fix multi-attach merged now and backported. | 17:20 |
| *** tejdeep has quit IRC | 17:21 | |
| jungleboyj | Once that is done, push up the rename. As long as we get that all done in the next couple of weeks it is fine. | 17:21 |
| *** tejdeep has joined #openstack-cinder | 17:21 | |
| pots | ok, that works for me. | 17:23 |
| *** enriquetaso has quit IRC | 17:23 | |
| jungleboyj | pots: Thank you! | 17:23 |
| *** _erlon_ has joined #openstack-cinder | 17:25 | |
| pots | thank you all for your patience | 17:25 |
| jungleboyj | pots: Thanks for keeping us in the loop as to what is going on. | 17:27 |
| *** enriquetaso has joined #openstack-cinder | 17:31 | |
| *** e0ne has quit IRC | 17:33 | |
| *** senrique_ has joined #openstack-cinder | 17:35 | |
| *** enriquetaso has quit IRC | 17:37 | |
| *** _hemna has joined #openstack-cinder | 17:49 | |
| geguileo | hemna: I figured out a solution to the problem, we just need to create a new iSCSI interface with a different initiator name :-) | 17:50 |
| geguileo | hemna: https://gorka.eguileor.com/host-iscsi-devices | 17:51 |
| hemna | checking | 18:13 |
| *** ociuhandu has joined #openstack-cinder | 18:16 | |
| hemna | so we get iscsi session isolation by creating a separate initiator interface and initiator iqn | 18:17 |
| hemna | which can be used by the host attaches | 18:17 |
| hemna | and then cinder/nova/os-brick attaches will use the default iqn | 18:17 |
| *** psachin has quit IRC | 18:19 | |
| hemna | we should document this as part of the cinder documentation | 18:20 |
| hemna | I can see others running into this | 18:20 |
| *** _hemna has quit IRC | 18:22 | |
| *** ociuhandu has quit IRC | 18:34 | |
| *** senrique__ has joined #openstack-cinder | 18:34 | |
| *** senrique__ is now known as enriquetaso | 18:34 | |
| *** senrique_ has quit IRC | 18:38 | |
| *** e0ne has joined #openstack-cinder | 18:42 | |
| *** mriedem has quit IRC | 18:42 | |
| *** enriquetaso has quit IRC | 18:49 | |
| *** ociuhandu has joined #openstack-cinder | 19:04 | |
| *** ociuhandu_ has joined #openstack-cinder | 19:09 | |
| *** ociuhandu has quit IRC | 19:09 | |
| *** whoami-rajat has quit IRC | 19:22 | |
| *** enriquetaso has joined #openstack-cinder | 19:35 | |
| *** henriqueof has quit IRC | 19:40 | |
| *** tosky has quit IRC | 19:47 | |
| *** mriedem has joined #openstack-cinder | 20:00 | |
| *** obi12341 has quit IRC | 20:02 | |
| *** jsquare has quit IRC | 20:03 | |
| *** jsquare has joined #openstack-cinder | 20:05 | |
| *** e0ne has quit IRC | 20:06 | |
| openstackgerrit | Pawel Kaminski proposed openstack/os-brick master: connectors/nvme: Wait utill nvme device show up in kernel https://review.opendev.org/672031 | 20:11 |
| *** _hemna has joined #openstack-cinder | 20:19 | |
| *** deiter has joined #openstack-cinder | 20:25 | |
| *** ociuhandu_ has quit IRC | 20:32 | |
| *** ociuhandu has joined #openstack-cinder | 20:33 | |
| *** pcaruana has quit IRC | 20:50 | |
| *** _hemna has quit IRC | 20:53 | |
| *** ociuhandu has quit IRC | 21:02 | |
| *** jrubenst has quit IRC | 21:15 | |
| *** gnufied has quit IRC | 21:21 | |
| *** lpetrut has joined #openstack-cinder | 21:22 | |
| *** e0ne has joined #openstack-cinder | 21:22 | |
| *** lpetrut has quit IRC | 21:22 | |
| *** lpetrut has joined #openstack-cinder | 21:23 | |
| *** lpetrut has quit IRC | 21:30 | |
| *** irclogbot_2 has quit IRC | 21:32 | |
| *** altlogbot_0 has quit IRC | 21:33 | |
| *** altlogbot_0 has joined #openstack-cinder | 21:33 | |
| *** irclogbot_2 has joined #openstack-cinder | 21:33 | |
| *** rosmaita has left #openstack-cinder | 21:39 | |
| *** enriquetaso has quit IRC | 21:44 | |
| *** e0ne has quit IRC | 21:50 | |
| *** gnufied has joined #openstack-cinder | 21:58 | |
| *** irclogbot_2 has quit IRC | 21:59 | |
| *** altlogbot_0 has quit IRC | 22:01 | |
| *** altlogbot_1 has joined #openstack-cinder | 22:21 | |
| *** TxGirlGeek has quit IRC | 22:25 | |
| *** altlogbot_1 has quit IRC | 22:27 | |
| *** _hemna has joined #openstack-cinder | 22:49 | |
| *** tkajinam has joined #openstack-cinder | 22:51 | |
| *** mriedem has quit IRC | 22:55 | |
| *** altlogbot_0 has joined #openstack-cinder | 23:13 | |
| *** carloss has quit IRC | 23:16 | |
| *** rcernin has joined #openstack-cinder | 23:16 | |
| *** altlogbot_0 has quit IRC | 23:19 | |
| *** _hemna has quit IRC | 23:24 | |
| *** altlogbot_0 has joined #openstack-cinder | 23:27 | |
| *** irclogbot_3 has joined #openstack-cinder | 23:31 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!