| *** ddieterly is now known as ddieterly[away] | 01:05 | |
| *** ddieterly[away] is now known as ddieterly | 01:40 | |
| *** ddieterly is now known as ddieterly[away] | 01:45 | |
| *** dschroeder has quit IRC | 01:51 | |
| *** ddieterly[away] is now known as ddieterly | 01:56 | |
| *** ddieterly is now known as ddieterly[away] | 01:57 | |
| *** ddieterly[away] has quit IRC | 02:57 | |
| *** c00281451 has joined #openstack-freezer | 03:25 | |
| *** dmellado_ has joined #openstack-freezer | 03:31 | |
| *** chenzeng has quit IRC | 03:32 | |
| *** dmellado has quit IRC | 03:32 | |
| *** EinstCrazy has joined #openstack-freezer | 04:33 | |
| *** EinstCra_ has joined #openstack-freezer | 04:57 | |
| *** EinstCrazy has quit IRC | 05:01 | |
| openstackgerrit | Einst Crazy proposed openstack/freezer: Restore cinder native backup pass right params and call right api https://review.openstack.org/299277 | 06:09 |
|---|---|---|
| *** daemontool has joined #openstack-freezer | 06:22 | |
| daemontool | I'll be on line in ~1:15 hours, later | 06:59 |
| *** daemontool has quit IRC | 07:04 | |
| *** c00281451 is now known as zengchen | 07:21 | |
| *** reldan has joined #openstack-freezer | 07:42 | |
| *** reldan has quit IRC | 08:04 | |
| *** reldan has joined #openstack-freezer | 08:26 | |
| *** daemontool has joined #openstack-freezer | 08:41 | |
| daemontool | hi | 08:42 |
| *** daemontool has quit IRC | 08:48 | |
| *** vannif has quit IRC | 08:53 | |
| *** daemontool has joined #openstack-freezer | 08:58 | |
| *** vannif has joined #openstack-freezer | 08:59 | |
| dmellado_ | hi daemontool ;) | 09:11 |
| dmellado_ | nice logo! ;) | 09:11 |
| daemontool | ty, that is courtesy of HPE | 09:15 |
| daemontool | :) | 09:15 |
| daemontool | vannif, ping | 09:23 |
| *** daemontool has quit IRC | 09:43 | |
| *** daemontool has joined #openstack-freezer | 09:44 | |
| daemontool | all, about the identification of the backup engine used, from the backup data binary blob | 09:48 |
| daemontool | any feedback? | 09:48 |
| daemontool | so from a conversation with reldan we have the following options | 09:48 |
| daemontool | 1) get the incremental_engine from metadata (in this case we need to have also the metadata to restore) | 09:49 |
| daemontool | 2) write a small header at the beginning of the data blob (i.e. freezer_rsync_v.0.1) | 09:49 |
| daemontool | 3) prepend rsync_ or tar_ at the beginning of the backup data blob | 09:50 |
| reldan | daemontool: I would prefer to store it in path like engine/backup/timestamp ... | 09:55 |
| reldan | But for this user should provide rsync or tar in command line | 09:55 |
| reldan | like restore —engine rsync | 09:55 |
| daemontool | slashme, vannif m3m0 all, ^^ ? | 09:56 |
| openstackgerrit | Merged openstack/freezer: Recreation swift client for each chunk during swift incremental restore https://review.openstack.org/298817 | 10:01 |
| daemontool | yes, 4) set --engine or --incremental-engine to the restore | 10:04 |
| reldan | daemontool: Here also may be a problem for incremental backup | 10:17 |
| reldan | daemontool: For incremental backup we need to find the latest backup by the same engine | 10:17 |
| reldan | daemontool: And if we have 2 - we should download backup by backup to find the latest with the same engine | 10:17 |
| reldan | daemontool: for 1 we should download metadata files one by one and check engine in metadata it should be the same | 10:18 |
| daemontool | ok | 10:20 |
| daemontool | what about number 3? | 10:20 |
| *** EinstCra_ has quit IRC | 10:21 | |
| reldan | daemontool: number 3 is good. I would prefer to instead of engine_hostname_backupname_timestamp_level have engine/hostname_backupname/timestamp/level | 10:21 |
| reldan | or engine/hostname/backupname/timestamp/level | 10:21 |
| daemontool | that would be the case for ssh and local fs right? | 10:21 |
| reldan | and for swift - swift supports sub-containers | 10:21 |
| reldan | pseudo-containers | 10:22 |
| reldan | so swift,local,ssh will have common structure | 10:22 |
| daemontool | there are quite few things to change | 10:23 |
| daemontool | which is not an issue | 10:23 |
| daemontool | so you want to do that change | 10:23 |
| daemontool | before rsync? | 10:23 |
| daemontool | by you I mean we | 10:23 |
| reldan | I don’t know. Because this change makes old backups obsolete | 10:24 |
| reldan | https://blueprints.launchpad.net/freezer/+spec/freezer-metadata | 10:24 |
| reldan | I can do, but it brings you additional problems during merge | 10:24 |
| reldan | So if you can do it with prefix, let’s say - probably we can try to do it with prefix. But I can make change it really quickly, but our old backups will be unsupported by new code | 10:25 |
| *** dmellado_ is now known as dmellado | 10:52 | |
| *** reldan has quit IRC | 11:10 | |
| *** dhellmann has quit IRC | 11:22 | |
| *** dhellmann_ has joined #openstack-freezer | 11:22 | |
| *** dhellmann_ is now known as dhellmann | 11:26 | |
| openstackgerrit | Merged openstack/freezer: Allow the user to use percentage on lvm snapsize https://review.openstack.org/298745 | 11:33 |
| daemontool | in newton I think we'll be backwards incompatible anyway | 11:40 |
| daemontool | can't remember the specific reason | 11:40 |
| daemontool | but I remember there was a discussion | 11:40 |
| daemontool | at the meetup | 11:40 |
| daemontool | where this was one of the agreement, if I remember well | 11:41 |
| *** reldan has joined #openstack-freezer | 12:07 | |
| slashme | The agreement was that if it needs to happen because of one on the big needed change we want introduce, we agreed it could happen and be mitigated by any of the following: Migration tool, documentation on how to use an old version, keeping the old freezer (as deprecated) code for one version with the freezerc entrypoint and deleting it in ocata. | 12:12 |
| slashme | If I remember well. We also said that if any big change needed to happen in the api we could introduce /v2. But that should not be necessary. | 12:14 |
| *** EinstCrazy has joined #openstack-freezer | 12:19 | |
| *** daemontool_ has joined #openstack-freezer | 12:39 | |
| *** daemontool has quit IRC | 12:42 | |
| *** daemontool__ has joined #openstack-freezer | 12:50 | |
| *** dmellado is now known as dmellado|lunch | 12:50 | |
| *** daemontool_ has quit IRC | 12:52 | |
| *** daemontool__ has quit IRC | 12:55 | |
| m3m0 | guys, let's move to #openstack-meeting-alt | 12:56 |
| *** daemontool__ has joined #openstack-freezer | 13:00 | |
| *** daemontool__ is now known as daemontool | 13:00 | |
| m3m0 | the meeting will be moved at 15:00 UTC | 13:01 |
| daemontool | m3m0, why? | 13:01 |
| daemontool | shouldn't be now? | 13:02 |
| m3m0 | the meeting boot is UTC and we set the meeting in GMT | 13:02 |
| m3m0 | so we need to wait one more hour | 13:02 |
| daemontool | http://eavesdrop.openstack.org/#Freezer_Meeting | 13:02 |
| daemontool | 1400 UTC | 13:03 |
| daemontool | ok | 13:03 |
| m3m0 | slashme is fixing that | 13:04 |
| daemontool | where do we set the meeting in GMT? | 13:04 |
| slashme | Wiki | 13:04 |
| slashme | Updated to put the right UTC | 13:04 |
| daemontool | ok | 13:05 |
| m3m0 | daemontool have you test how big the metadata gets for rsync when a backup has a million files? | 13:14 |
| *** szaher_ has joined #openstack-freezer | 13:24 | |
| *** yangyapeng has joined #openstack-freezer | 13:24 | |
| daemontool | is depends on how files are big | 13:26 |
| daemontool | for the API is not an issue as we do not upload the data to the api | 13:26 |
| daemontool | the metadata is compressed | 13:27 |
| daemontool | I mean the one in the storage media | 13:27 |
| *** ddieterly has joined #openstack-freezer | 13:32 | |
| *** dmellado|lunch is now known as dmellado | 13:54 | |
| reldan | meeting? | 14:01 |
| m3m0 | si | 14:01 |
| m3m0 | guys, let's move to #openstack-meeting-alt | 14:03 |
| yangyapeng | :） | 14:03 |
| *** szaher_ has quit IRC | 14:04 | |
| *** ddieterly is now known as ddieterly[away] | 14:56 | |
| *** ddieterly[away] is now known as ddieterly | 15:03 | |
| ddieterly | slashme ping | 15:06 |
| *** ddieterly is now known as ddieterly[away] | 15:17 | |
| *** dschroeder has joined #openstack-freezer | 15:18 | |
| *** ddieterly[away] is now known as ddieterly | 15:38 | |
| *** EinstCrazy has quit IRC | 15:47 | |
| *** ddieterly is now known as ddieterly[away] | 16:16 | |
| *** ddieterly[away] is now known as ddieterly | 16:19 | |
| openstackgerrit | Eldar Nugaev proposed openstack/freezer: Fix glance client creation https://review.openstack.org/300075 | 16:22 |
| openstackgerrit | Cynthia Lopes do Sacramento proposed openstack/freezer-api: Add consistency_checksum parameter on backup metadata https://review.openstack.org/300070 | 16:25 |
| *** ddieterly is now known as ddieterly[away] | 16:26 | |
| *** ddieterly[away] is now known as ddieterly | 16:29 | |
| openstackgerrit | Cynthia Lopes do Sacramento proposed openstack/freezer: Adds checksum before backup and after restore for consistency check. https://review.openstack.org/300080 | 16:30 |
| *** reldan has quit IRC | 16:33 | |
| openstackgerrit | Deklan Dieterly proposed openstack/freezer-api: Add Freezer API Tempest Tests https://review.openstack.org/297360 | 17:15 |
| *** ddieterly is now known as ddieterly[away] | 17:23 | |
| *** daemontool has quit IRC | 17:25 | |
| *** ddieterly[away] has quit IRC | 18:23 | |
| *** reldan has joined #openstack-freezer | 18:41 | |
| *** reldan has quit IRC | 18:55 | |
| *** openstackgerrit has quit IRC | 19:03 | |
| *** openstackgerrit has joined #openstack-freezer | 19:04 | |
| *** ddieterly has joined #openstack-freezer | 19:13 | |
| *** ddieterly has quit IRC | 19:17 | |
| *** reldan has joined #openstack-freezer | 19:27 | |
| *** reldan has quit IRC | 19:55 | |
| *** reldan has joined #openstack-freezer | 19:56 | |
| *** reldan has quit IRC | 20:53 | |
| *** dschroeder has quit IRC | 22:25 | |
| *** reldan has joined #openstack-freezer | 22:39 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!