*** haomaiwa_ has quit IRC | 00:01 | |
*** haomaiwang has joined #openstack-manila | 00:01 | |
openstackgerrit | Merged openstack/manila: Admin networks in NetApp cDOT multi-SVM driver https://review.openstack.org/281950 | 00:32 |
---|---|---|
*** pgbridge has joined #openstack-manila | 00:35 | |
cknight | ganso: ping | 00:55 |
ganso | cknight: pong | 00:55 |
cknight | ganso: were you going to rebase your migration patches? | 00:55 |
ganso | cknight: yes. I am going to. | 00:55 |
cknight | ganso: I have the migration code running, but they don't include my driver's update_access code. | 00:56 |
cknight | ganso: I can cherry-pick that for now, but it'll help going forward if you can get my latest. | 00:56 |
ganso | cknight: sure. Cherry-picking should work right now. | 00:57 |
cknight | ganso: ok, will try. | 00:57 |
*** dalgaaf has quit IRC | 00:58 | |
*** haomaiwang has quit IRC | 01:01 | |
*** haomaiwang has joined #openstack-manila | 01:01 | |
cknight | ganso: just witnessed migration of an empty share on my driver | 01:10 |
cknight | ganso: now we'll try with some data in it | 01:10 |
ganso | cknight: =D | 01:10 |
cknight | ganso: (or do you not support migration of shares with data ;-)) | 01:10 |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Implement update_access() in generic driver + LVM https://review.openstack.org/256281 | 01:15 |
cknight | ganso: If I have access rules in place before a migration, what rules should be present after a migration? | 01:15 |
cknight | ganso: Because I'm seeing no access rules remaining after the migration. | 01:16 |
ganso | cknight: the same ones | 01:16 |
ganso | cknight: o_O | 01:16 |
ganso | cknight: do you have the rules in DB after migration? | 01:17 |
cknight | ganso: Manila says the migration succeeded, it shows on the other pool, I see the share on the other backend, but the access rules are gone. | 01:17 |
cknight | ganso: I didn't check the DB, but I figured Manila did. | 01:17 |
cknight | ganso: I can look in the DB if that would help. | 01:17 |
ganso | cknight: Migration code does not erase them from DB. After share is migrated, it calls update_access on new instance, applying all the access rules straight from DB | 01:18 |
openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs.common: move the numreduct function to toplevel https://review.openstack.org/282069 | 01:19 |
openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs_native: relocate module under glusterfs https://review.openstack.org/280935 | 01:19 |
openstackgerrit | Csaba Henk proposed openstack/manila: gluster*: clean up volume option querying https://review.openstack.org/277141 | 01:19 |
ganso | cknight: migration triggers update_access in recovery mode several times. The driver must support recovery mode | 01:19 |
openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs-native: use Heketi layout via self-signing https://review.openstack.org/282101 | 01:20 |
openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: heketi: Add support for JWT Auth https://review.openstack.org/280486 | 01:20 |
openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: introducing Heketi layout https://review.openstack.org/279090 | 01:20 |
cknight | ganso: because of how my backend works, I never use add_rules or delete_rules. I always replace everything using access_rules. | 01:20 |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Implement admin network in generic driver https://review.openstack.org/278699 | 01:21 |
cknight | ganso: all the rules in the DB show deleted. | 01:21 |
ganso | cknight: exactly, recovery mode does not use add_rules and delete_rules | 01:21 |
ganso | cknight: weird, I will re-test on rebase, they should never be deleted | 01:22 |
cknight | ganso: ok | 01:22 |
ganso | cknight: I am rebasing right now | 01:22 |
cknight | ganso: in the meantime, I can start reading the code. I had some initial feedback for you. | 01:22 |
openstackgerrit | Csaba Henk proposed openstack/manila: ganesha: implement update_access https://review.openstack.org/282602 | 01:22 |
ganso | cknight: great | 01:22 |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Introduced Data Service https://review.openstack.org/244286 | 01:27 |
*** haomaiwang has quit IRC | 01:30 | |
bswartz | ganso: ping | 01:40 |
ganso | bswartz: pong | 01:40 |
bswartz | so what's going on with the LVM share access methods | 01:40 |
bswartz | the in-tree driver still uses the legacy methods, and you patch fixes that? | 01:41 |
ganso | bswartz: the bug can easily be observable in CI job, because it is multibackend | 01:41 |
ganso | bswartz: NFS helper code has locks | 01:42 |
ganso | bswartz: but those locks are only per backend | 01:42 |
bswartz | which bug | 01:42 |
ganso | bswartz: concurrency | 01:42 |
bswartz | is the bug reported though? | 01:42 |
ganso | bswartz: one allow_access call overwrites the other | 01:42 |
ganso | bswartz: yes | 01:43 |
ganso | bswartz: it is one of those I showed you today | 01:43 |
bswartz | link? | 01:43 |
ganso | bswartz: https://bugs.launchpad.net/manila/+bug/1550378 | 01:43 |
openstack | Launchpad bug 1550378 in Manila "LVM access rules concurrency" [Medium,New] | 01:43 |
bswartz | oh lol | 01:43 |
bswartz | we just looked at this one this morning | 01:43 |
ganso | bswartz: if you create a debug patch that performs a "cat /etc/exports" you will be able to see the rules getting overwritten | 01:43 |
bswartz | yes | 01:44 |
bswartz | well let me show you what I'm seeing... | 01:44 |
bswartz | http://logs.openstack.org/29/285429/1/check/gate-manila-tempest-dsvm-neutron-postgres-lvm-multibackend/daed4b5/console.html | 01:45 |
bswartz | http://logs.openstack.org/29/285429/1/check/gate-manila-tempest-dsvm-neutron-postgres-lvm-multibackend/daed4b5/logs/screen-m-shr.txt.gz | 01:45 |
bswartz | http://paste.openstack.org/show/488445/ | 01:46 |
bswartz | ganso: ^ is that what you've observed? | 01:46 |
*** qeelee has joined #openstack-manila | 01:46 | |
ganso | bswartz: no, that is a different problem. | 01:46 |
ganso | bswartz: it may be related to this https://bugs.launchpad.net/manila/+bug/1550295 | 01:47 |
openstack | Launchpad bug 1550295 in Manila "Concurrent requests to allow access get lost" [High,New] - Assigned to Tiago Pasqualini da Silva (tiago.pasqualini) | 01:47 |
ganso | bswartz: the one you observed is concurrency in API layer | 01:47 |
ganso | bswartz: the one I am fixing is concurrency in LVM driver layer | 01:47 |
ganso | bswartz: in fact, lemme just confirm that for you | 01:48 |
bswartz | how can we get these fixed? | 01:48 |
bswartz | they're going to cause random gate test pain I think | 01:48 |
ganso | bswartz: cause, if that "ShareAccessExists" exception you observed came from the driver layer, then my patch should be fixing that as well | 01:48 |
ganso | bswartz: indeed, it comes from the driver layer | 01:49 |
ganso | bswartz: so, my patch may fix that | 01:49 |
bswartz | okay next question | 01:49 |
ganso | bswartz: if you could test on top of https://review.openstack.org/#/c/256281, maybe it could be fixed | 01:50 |
bswartz | this patch: https://review.openstack.org/#/c/256281 | 01:50 |
bswartz | it seems to use common code for generic and LVM | 01:50 |
bswartz | however it invokes SSH | 01:50 |
bswartz | SSH isn't needed for LVM | 01:50 |
bswartz | am I missing something? | 01:50 |
ganso | bswartz: just a sec | 01:51 |
*** mtanino_ has joined #openstack-manila | 01:51 | |
ganso | bswartz: yes, it does not need SSH | 01:51 |
ganso | bswartz: I believe that can be easily improved by passing a regular executor instead of a ssh executor when instantiating the helper | 01:52 |
ganso | bswartz: it already uses whichever executor is supplied in the constructor | 01:52 |
*** mtanino has quit IRC | 01:52 | |
bswartz | ah okay | 01:52 |
bswartz | ganso: would the concurrency bug also occur with a tempest concurrency of 1? | 01:56 |
bswartz | beacuse it's not related to a real concurrency issue, it's related to making back to back calls without waiting? | 01:56 |
ganso | bswartz: the one in the driver layer no. I tested it myself. | 01:56 |
bswartz | ok | 01:56 |
bswartz | that makes it harder to reproduce | 01:57 |
*** mtanino_ has quit IRC | 01:57 | |
cknight | ganso: still there? | 02:08 |
ganso | cknight: yes | 02:08 |
cknight | ganso: ok, I ran a couple experiments | 02:09 |
cknight | ganso: earlier I migrated an empty share, so first I placed a single test file on the share to be migrated. that failed during the data copy. I think it's because no access rules were added to the destination. | 02:09 |
cknight | ganso: I'm running my driver in a debugger, and I can see the access rules being set to read-only on the source, and then removed from the source. | 02:11 |
cknight | ganso: But I never see them set on the destination. | 02:11 |
ganso | cknight: one problem at a time. The data copy failed because it failed to mount, correct? | 02:11 |
cknight | ganso: Not sure whether the mount worked or not. I still need to check that. | 02:12 |
ganso | cknight: the only rule needed to be added to the destination for data copy is the data_node_access_ip | 02:14 |
ganso | cknight: and there must be connectivity between the data node and both share servers | 02:14 |
ganso | cknight: about the missing rules in the destination: the method "apply_new_access_rules" is invoked in migration_complete method (in manager) to apply the rules to the destination | 02:17 |
cknight | ganso: yes, I set that to my devstack system address, and I can mount shares from both backends to that address | 02:17 |
cknight | ganso: oh, I didn't run migration_complete. | 02:18 |
ganso | cknight: you don't have to | 02:18 |
cknight | ganso: i guess I don't really know how to use the feature yet | 02:18 |
ganso | cknight: you only have to run migration_complete if you migrate a share with --notify False | 02:18 |
bswartz | errr | 02:19 |
cknight | ganso: ok, perhaps we should focus initially on the empty share case, since it appears to succeed but doesn't copy the rules | 02:19 |
bswartz | I just realized I don't know how to run tempest now since we moved the test to the plugin directory | 02:19 |
cknight | bswartz: +1 | 02:19 |
bswartz | does anyone know the new command? | 02:19 |
ganso | bswartz: I do, hold on | 02:19 |
cknight | bswartz: to run Tempest, I use 'git review' | 02:20 |
ganso | bswartz: tox -eall-plugin manila_tempest_tests.tests.api.admin.test_migration -- --concurrency=1 | tee teste2.log | 02:20 |
ganso | bswartz: don't forget to rollback tempest to compatible commit | 02:20 |
bswartz | ganso: what directories to you run that command in? | 02:23 |
bswartz | in the tempest dir or the manila dir | 02:24 |
ganso | bswartz: tempest dir | 02:24 |
bswartz | ganso: do you copy any files before running that? | 02:26 |
ganso | bswartz: nope | 02:26 |
bswartz | I don't understand how tempest magically finds the manila tests | 02:26 |
*** qeelee has quit IRC | 02:28 | |
*** qeelee has joined #openstack-manila | 02:31 | |
* bswartz curses himself | 02:31 | |
bswartz | GIT_DEPTH screws me with the tempest commit | 02:31 |
cknight | ganso: I set a breakpoint at data/manager/migration_start. At that point, I can see the destination share was created, but no access rules exist there and it isn't mounted. I don't see any errors in any of the logs. | 02:32 |
ganso | cknight: so far, it is correct | 02:32 |
ganso | cknight: in the beginning of data/manager/migration_start, destination share should not have any access rules, it is added in that data/manager/migration_start method. | 02:33 |
cknight | ganso: ok, good. I'll keep stepping through, then. | 02:33 |
ganso | cknight: original rules remain only in source share, in RO mode | 02:34 |
*** gouthamr has joined #openstack-manila | 02:36 | |
*** qeelee has quit IRC | 02:40 | |
*** qeelee has joined #openstack-manila | 02:41 | |
cknight | ganso: [Errno 13] Permission denied: u'/tmp/57907eca-3ef7-4a06-9cf4-003aabb11a6c/testfile' | 02:45 |
cknight | ganso: I see a rw rule at the destination. | 02:46 |
cknight | ganso: And a rw rule at the source. | 02:46 |
ganso | cknight: that was a permission denied to mount or to copy? did you create a file with strict permission? | 02:46 |
cknight | ganso: Both source & dest are mounted. | 02:46 |
ganso | cknight: is that rw rule the same one as data_node_access_ip? | 02:47 |
ganso | cknight: if yes, it is correct | 02:47 |
cknight | ganso: yes, it is data_node_access_ip | 02:47 |
cknight | ganso: not sure why the source is rw, but not a big deal | 02:47 |
cknight | ganso: and yes, the test file is: -rw-r--r-- 1 nobody 4294967294 21 May 2 2015 testfile | 02:48 |
cknight | ganso: is that a problem? | 02:48 |
ganso | cknight: it is to prevent additional logic to address backends that do not support RO mode, like CIFS IP | 02:48 |
cknight | ganso: makes sense, thanks | 02:48 |
ganso | cknight: I have not tried to address cases like that, files without permission to access | 02:49 |
ganso | cknight: but probably copying as root with address that | 02:49 |
ganso | cknight: my idea was that the admin knows what he is copying, since migration is admin_only | 02:50 |
ganso | cknight: so he should have permissions sorted out | 02:50 |
cknight | ganso: But the admin is migrating user data, which could have any permissions. | 02:50 |
cknight | ganso: I think you'll need root permission to copy everything. | 02:50 |
ganso | cknight: I will take a look at shutil to see if it supports root | 02:51 |
cknight | ganso: I left a comment about the copy mechanism you're using. I wonder if 'cp -a' with rootwrap would work better. It would also preserve the ownership & permissions. | 02:51 |
ganso | cknight: I haven't replied to that one yet | 02:51 |
ganso | cknight: I tried using rsync before and I would get I/O errors | 02:51 |
ganso | cknight: random I/O errors | 02:52 |
cknight | ganso: Probably want bswartz to weigh in. I'll try a simpler test before heading home. | 02:52 |
ganso | cknight: I tried 'cp -a' before trying rsync, I don't remember the problems I faced | 02:52 |
ganso | cknight: but since I changed everything to copy file by file with shutil, it would probably be easier to switch to 'cp -a' single files right now | 02:52 |
ganso | cknight: so, regarding access rules, so far so good in your debugging | 02:53 |
*** vbellur has joined #openstack-manila | 02:55 | |
* bswartz is neck deep in tempest | 02:55 | |
bswartz | sorry I'm not reading the channel | 02:55 |
bswartz | anything I can answer? | 02:55 |
cknight | bswartz: not really, just chatting with ganso about the file copy mechanism | 02:56 |
bswartz | cp -a seems reasonable to me | 02:56 |
*** yangyapeng has quit IRC | 03:04 | |
*** yangyapeng has joined #openstack-manila | 03:04 | |
cknight | bswartz: How do I fix this? umount.nfs: /mnt/manila: device is busy | 03:06 |
bswartz | cknight: "lsof /mnt/manila" | 03:07 |
cknight | clinton@mitaka:~$ lsof /mnt/manila/ | 03:07 |
cknight | COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME | 03:07 |
cknight | ls 75599 clinton cwd DIR 0,47 4096 64 /mnt/manila (192.168.228.110:/share_dc01def1_473b_49b5_8a98_8aea3bed41a3) | 03:07 |
cknight | ls 75599 clinton 3r DIR 0,47 4096 64 /mnt/manila (192.168.228.110:/share_dc01def1_473b_49b5_8a98_8aea3bed41a3) | 03:07 |
bswartz | that lists the process still using the filesystem, and thus preventing the unmount | 03:08 |
bswartz | kill those processes and try the umount again | 03:08 |
cknight | thanks, that did it | 03:09 |
*** yangyapeng has quit IRC | 03:12 | |
bswartz | ganso: tempest is having all kind of auth issues | 03:19 |
bswartz | ganso: did you do any keystone things before running tempest? | 03:19 |
ganso | https://www.irccloud.com/pastebin/DTXQ2UAm/ | 03:20 |
ganso | bswartz: ^ | 03:20 |
bswartz | yeah I basically have that | 03:21 |
bswartz | oh wait | 03:21 |
bswartz | password could be wrong >_< | 03:21 |
bswartz | yeah it was the password in tempest.conf | 03:22 |
bswartz | thx | 03:22 |
bswartz | woohoo tempest running! | 03:22 |
bswartz | so it's still not clear to me how tempest find the path to the manila plugin | 03:22 |
bswartz | fins | 03:22 |
bswartz | finds | 03:22 |
ganso | bswartz: magic | 03:26 |
cknight | bswartz: nfs4 was getting hung up, nfs3 working better | 03:26 |
cknight | ganso: copy failed again with a world-readable file | 03:26 |
bswartz | cknight: for NFS4 you need significantly more configuration for it not to blow up | 03:26 |
cknight | ganso: i think the dest directory isn't writable | 03:26 |
ganso | cknight: is that something particular of your backend? | 03:27 |
ganso | cknight: are you mounting using sudo? | 03:27 |
ganso | cknight: nevermind, code is | 03:27 |
cknight | ganso: right. this is the dest mount: drwxr-xr-x 2 root root 4096 Feb 26 22:22 6eaccb81-5fd7-44f8-9653-4108a2ca036b/ | 03:28 |
cknight | ganso: so you'll need root access to copy there | 03:28 |
ganso | cknight: wait, did you create the file as root? | 03:28 |
ganso | cknight: I usually use sudo to mount, then create files as user | 03:28 |
bswartz | cknight -- remember that NFS often has root_squash enabled | 03:29 |
bswartz | if we want to allow I/O as root we need to configure the export rules for root access | 03:29 |
cknight | ganso: I did a chown to myself and then chmod 666 | 03:29 |
bswartz | directory should be 777 | 03:29 |
cknight | ganso: here is the source dir: drwxrwxrwx 2 root root 4096 May 2 2015 a7b16079-dce3-4226-866f-c9490a078ef6/ | 03:30 |
ganso | cknight: it is created as root | 03:31 |
cknight | ganso: here is the source file: -rw-rw-rw- 1 clinton root 31 May 2 2015 testfile | 03:31 |
cknight | ganso: so what did I do wrong? | 03:31 |
ganso | cknight: it seems to me folder is created as root | 03:32 |
ganso | cknight: it should be clinton user | 03:32 |
cknight | ganso: those are just the nfs mounts | 03:32 |
bswartz | ganso: the directory will get a new uid/gid from the NFS server after the mount | 03:33 |
bswartz | netapp default to ownership of the NFS root directory by root user | 03:33 |
ganso | hummm | 03:33 |
cknight | bswartz: thanks, that makes a lot more sense | 03:33 |
bswartz | and cknight you'll love this -- we have no API to chown directories | 03:33 |
ganso | I will take a look into copying files as root, should address this | 03:34 |
cknight | So we need elevated permissions to do the copy? | 03:34 |
cknight | Yes. | 03:34 |
cknight | But I'm also concerned about your explore method. | 03:34 |
bswartz | ganso -- you'd need root permission to preserve uid/gid on the files you copied anyways | 03:34 |
cknight | It builds a list of files & dirs in memory. | 03:34 |
ganso | I have not faced these problems with generic or HDS backend | 03:35 |
cknight | It seems that's not gonna scale. | 03:35 |
*** qeelee has quit IRC | 03:35 | |
ganso | cknight: should be a big list of strings | 03:35 |
ganso | cknight: if we do not have many multiple copy operations at the same time, it should not use too much memory | 03:36 |
openstackgerrit | Merged openstack/manila: Change sudo to run_as_root in LVM driver https://review.openstack.org/285107 | 03:39 |
cknight | ganso: If 'cp -aR' won't work and we have to recurse the files in the data service, I think it would be better to copy each file as we go instead of building a potentially massive list. | 03:39 |
cknight | ganso: That way you aren't storing anything. | 03:40 |
ganso | cknight: cp -ar will probably not be able to track progress | 03:41 |
ganso | cknight: or even cancel | 03:41 |
cknight | ganso: OK, I'll buy that. | 03:41 |
cknight | ganso: So hopefully we combine explore and copy into a single method that doesn't store the whole file list. And use 'cp' to do the copy fast and to preserve the metadata. | 03:42 |
cknight | ganso: And you can add a rootwrap for 'cp' so it doesn't get stuck. | 03:42 |
ganso | cknight: and as sudo | 03:42 |
ganso | cknight: yes | 03:42 |
cknight | ganso: not sudo explicitly, use rootwrap | 03:43 |
ganso | cknight: run_as_root=True, right? | 03:43 |
cknight | ganso: think so, yes. | 03:43 |
bswartz | I disagree | 03:43 |
bswartz | the copy has to be done as root completely | 03:43 |
ganso | bswartz: yes | 03:44 |
bswartz | we can't be enumerating the file tree inside manila -- manila doesn't have root permission so we will fail to read any 700 directories | 03:44 |
cknight | bswartz: the code is doing that now | 03:44 |
bswartz | there are many ways to be copying files | 03:44 |
bswartz | cp -a is a good start | 03:44 |
bswartz | if the only problem is tracking progress, I can think of other approaches to address that | 03:45 |
bswartz | any kind of python code is a bad idea though | 03:45 |
bswartz | we don't want to reinvent this particular wheel | 03:46 |
cknight | bswartz: how to track progress? look at cumulative file sizes? | 03:46 |
bswartz | one old trick is "tar cf - /src | tar xf - /dst" | 03:47 |
ganso | cknight: how do I fix this? Scheduler host filter ShareReplicationFilter could not be found. | 03:47 |
bswartz | you can count the bytes that travel through the pipe | 03:47 |
*** furlongm_ has quit IRC | 03:49 | |
ganso | cknight: I remember running into this error before | 03:49 |
ganso | cknight: when consistency groups merged | 03:49 |
cknight | ganso: checking | 03:50 |
ganso | cknight: found it, entry_points.txt | 03:50 |
ganso | cknight: need to add the filter specifically somewhere | 03:50 |
cknight | ganso: yes, in entry_points.txt | 03:51 |
cknight | ganso: add to manila.scheduler.filters stanza | 03:51 |
ganso | cknight: got it | 03:52 |
ganso | cknight: I am testing access rules problem right now | 03:52 |
cknight | ganso: ok. after a failed migration, I don't see the rules restored on the source. | 03:53 |
cknight | ganso: I do see everything else cleaned up, though. | 03:53 |
ganso | cknight: rules on the source are empty? or RO? | 03:53 |
cknight | ganso: empty | 03:53 |
cknight | ganso: empty in Manila, and empty on the storage controller | 03:54 |
cknight | bswartz, ganso: well guys, I need to head home for dinner. I see 14 patches merged today, with another in the gate and several more +1's from Goutham. So I feel much better about the deadline than I did 2 days ago. | 03:55 |
cknight | ganso: thanks for working late on this. it seems most of the migration work ahead is the actual copy mechanism. bswartz is your best resource for help there. | 03:56 |
bswartz | dinner?!? | 03:56 |
bswartz | it's almost 11pm | 03:56 |
gouthamr | cknight: you mean breakfast in bed? ;) | 03:56 |
cknight | I know. I'm leaving now. But it's so nice & quiet here. | 03:57 |
ganso | cknight: good night cknight, thanks for the help and for testing migration | 03:57 |
cknight | ganso: np. have a great weekend, guys. | 03:58 |
bswartz | ganso: it must be past midnight where you are | 03:58 |
ganso | bswartz: 1am | 03:58 |
*** cknight has quit IRC | 03:58 | |
*** qeelee has joined #openstack-manila | 03:59 | |
gouthamr | ganso: where do you work out of? | 03:59 |
ganso | gouthamr: sorry I did not understand the question, you mean where I am based? | 03:59 |
bswartz | Sao Paolo | 03:59 |
bswartz | right? | 04:00 |
ganso | bswartz: s/Sao Paolo/Sao Paulo | 04:00 |
gouthamr | ganso: er. sorry, yes. | 04:00 |
bswartz | typo | 04:00 |
gouthamr | ganso: nice. | 04:00 |
ganso | gouthamr: you? | 04:00 |
gouthamr | ganso: Raleigh in NC.. | 04:01 |
gouthamr | ganso: a couple cubes away from bswartz | 04:01 |
bswartz | ganso: all of us from NetApp are in Raleigh NC | 04:01 |
ganso | bswartz, gouthamr: are you guys still getting a lot of snow? | 04:01 |
bswartz | ganso: not currently | 04:02 |
bswartz | it's been high teens and rainy | 04:02 |
bswartz | nothing below freezing last 2 weeks | 04:02 |
gouthamr | ganso bswartz: which is nice 8) | 04:03 |
gouthamr | vponomaryov: are you here? | 04:05 |
bswartz | it's saturday in ukraine | 04:05 |
bswartz | any still 6am | 04:05 |
bswartz | and still 6am | 04:05 |
*** martyturner has joined #openstack-manila | 04:05 | |
gouthamr | bswartz: yeah, wishful thinking.. i wasn't able to get that experimental job to run on the zfsonlinux driver | 04:05 |
bswartz | you couldn't get it to pass or you couldn't get it to run? | 04:06 |
ganso | I will have some bad news for cknight, I just tested migration with generic driver and access rules are fine. Something weird going on with NetApp drivers | 04:06 |
ganso | with NetApp + migration | 04:06 |
bswartz | gouthamr: I've gotten tempest running now on my dev system thanks to help from ganso | 04:08 |
gouthamr | bswartz: I'm making an edit in the post_test_hook of vponomaryov's patch | 04:08 |
bswartz | and ZFS is trivial to get running, so if you have some patches you want me to cherrypick and run, just give them the numbers | 04:08 |
gouthamr | i'm hoping that'll do it | 04:08 |
bswartz | k | 04:08 |
bswartz | gouthamr: so you want it to run in infra cloud not locally? | 04:09 |
gouthamr | bswartz: yep.. he has an experimental pipeline; thought i'd use it | 04:09 |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Move Share Migration code to Data Service https://review.openstack.org/250515 | 04:10 |
*** martyturner has quit IRC | 04:10 | |
*** qeelee has quit IRC | 04:10 | |
*** gouthamr_ has joined #openstack-manila | 04:10 | |
openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 04:10 |
*** qeelee has joined #openstack-manila | 04:11 | |
*** qeelee has quit IRC | 04:12 | |
*** yangyapeng has joined #openstack-manila | 04:12 | |
ganso | bswartz, gouthamr: going to bed now guys | 04:12 |
bswartz | goodnight | 04:13 |
gouthamr_ | goodnight ganso! | 04:13 |
ganso | bswartz, gouthamr: have a nice weekend =) | 04:13 |
gouthamr_ | ganso: you too :) | 04:13 |
ganso | bswartz, gouthamr: thanks! good night! | 04:13 |
*** ganso has left #openstack-manila | 04:13 | |
*** gouthamr has quit IRC | 04:13 | |
*** jasonsb has joined #openstack-manila | 04:14 | |
openstackgerrit | Merged openstack/manila: gluster*: clean up volume option querying https://review.openstack.org/277141 | 04:15 |
*** pgbridge has quit IRC | 04:16 | |
openstackgerrit | Ben Swartzlander proposed openstack/manila: Add doc for export location metadata https://review.openstack.org/282071 | 04:17 |
*** yangyapeng has quit IRC | 04:18 | |
openstackgerrit | Merged openstack/manila: Improve exception msg when attaching/detaching volumes https://review.openstack.org/282220 | 04:38 |
*** zhonghua has quit IRC | 04:45 | |
*** zhonghua-lee has joined #openstack-manila | 04:46 | |
openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 04:47 |
openstackgerrit | Goutham Pacha Ravi proposed openstack/python-manilaclient: Client Support for Manila Share Replication https://review.openstack.org/235448 | 05:00 |
*** sgotliv has joined #openstack-manila | 05:00 | |
*** vbellur has quit IRC | 05:04 | |
*** yangyapeng has joined #openstack-manila | 05:18 | |
*** yangyapeng has quit IRC | 05:23 | |
*** gouthamr_ has quit IRC | 05:26 | |
*** haomaiwa_ has joined #openstack-manila | 05:32 | |
*** haomaiwa_ has quit IRC | 06:01 | |
*** haomaiwang has joined #openstack-manila | 06:01 | |
*** bogdans has joined #openstack-manila | 06:47 | |
*** zhonghua-lee has quit IRC | 06:56 | |
*** zhonghua-lee has joined #openstack-manila | 06:56 | |
*** haomaiwang has quit IRC | 07:01 | |
*** haomaiwang has joined #openstack-manila | 07:01 | |
*** bogdans has quit IRC | 07:06 | |
*** yangyapeng has joined #openstack-manila | 07:25 | |
*** yangyapeng has quit IRC | 07:33 | |
*** sgotliv has quit IRC | 07:35 | |
openstackgerrit | liucheng proposed openstack/manila: Update Huawei driver doc for Mitaka https://review.openstack.org/284581 | 07:57 |
*** haomaiwang has quit IRC | 08:01 | |
*** haomaiwang has joined #openstack-manila | 08:01 | |
openstackgerrit | zhaohua proposed openstack/manila: Implement update_access() method in huawei driver https://review.openstack.org/266310 | 08:18 |
*** yangyapeng has joined #openstack-manila | 08:32 | |
*** yangyapeng has quit IRC | 08:38 | |
*** haomaiwang has quit IRC | 09:01 | |
*** haomaiwang has joined #openstack-manila | 09:01 | |
*** lpetrut has joined #openstack-manila | 09:37 | |
*** haomaiw__ has joined #openstack-manila | 09:45 | |
*** haomaiwang has quit IRC | 09:46 | |
*** haomaiwang has joined #openstack-manila | 09:48 | |
*** haomaiw__ has quit IRC | 09:49 | |
*** sgotliv has joined #openstack-manila | 09:52 | |
*** haomaiwang has quit IRC | 10:01 | |
*** haomaiwang has joined #openstack-manila | 10:01 | |
*** haomaiwang has quit IRC | 10:17 | |
*** yangyapeng has joined #openstack-manila | 10:40 | |
*** yangyapeng has quit IRC | 10:48 | |
*** lpetrut has quit IRC | 10:50 | |
*** lpetrut has joined #openstack-manila | 10:54 | |
*** haomaiwa_ has joined #openstack-manila | 11:00 | |
*** haomaiwa_ has quit IRC | 11:01 | |
*** haomaiwang has joined #openstack-manila | 11:02 | |
*** openstack has joined #openstack-manila | 23:32 | |
*** openstackstatus has quit IRC | 23:36 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!