Monday, 2013-12-16

« Sunday, 2013-12-15 Index Tuesday, 2013-12-17 »
*** rongze has joined #openstack-trove00:31
*** rongze has quit IRC00:37
*** mmcdaris has joined #openstack-trove00:55
*** mmcdaris has quit IRC00:58
*** nosnos has joined #openstack-trove01:05
*** freyes has joined #openstack-trove01:41
*** erkules has quit IRC02:19
*** erkules_ has joined #openstack-trove02:19
*** rongze has joined #openstack-trove02:23
*** rongze_ has joined #openstack-trove02:31
*** rongze has quit IRC02:33
*** robertmyers has joined #openstack-trove03:01
*** _Seal has joined #openstack-trove03:20
*** yogesh has joined #openstack-trove03:26
*** yogesh has quit IRC03:31
*** yogesh has joined #openstack-trove03:32
*** yogesh has quit IRC03:36
openstackgerritSushil Kumar proposed a change to openstack/trove: Removes dead code from trove/tests/api/root.py  https://review.openstack.org/6223603:39
*** freyes has quit IRC03:41
*** shalini has joined #openstack-trove03:42
*** jcooley_ has quit IRC03:51
*** _Seal has quit IRC04:02
*** mrsnivvel has joined #openstack-trove04:23
*** robertmyers has quit IRC04:43
*** rongze_ has quit IRC04:44
*** jcooley_ has joined #openstack-trove04:46
*** haomaiwang has joined #openstack-trove05:09
*** SergeyLukjanov has joined #openstack-trove05:11
*** haomaiwang has quit IRC05:14
*** haomaiwang has joined #openstack-trove05:18
*** jcooley_ has quit IRC05:19
*** haomaiwang has quit IRC05:23
*** nosnos_ has joined #openstack-trove05:34
*** rongze has joined #openstack-trove05:35
*** nosnos_ has quit IRC05:35
*** nosnos_ has joined #openstack-trove05:36
*** nosnos has quit IRC05:37
*** rongze has quit IRC05:41
*** erkules_ is now known as erkules05:41
*** haomaiwang has joined #openstack-trove05:49
*** jcooley_ has joined #openstack-trove06:01
*** yogesh_ has joined #openstack-trove06:01
*** rongze has joined #openstack-trove06:08
*** SergeyLukjanov is now known as _SergeyLukjanov06:11
*** jcooley_ has quit IRC06:20
*** _SergeyLukjanov is now known as SergeyLukjanov06:20
*** yogesh_ has quit IRC06:21
*** jcooley_ has joined #openstack-trove06:36
*** yogesh_ has joined #openstack-trove06:37
*** jcooley_ has quit IRC06:39
*** _Seal has joined #openstack-trove06:44
*** SushilKM has joined #openstack-trove06:44
*** nosnos has joined #openstack-trove06:59
*** nosnos_ has quit IRC06:59
*** SushilKM has quit IRC07:44
*** SushilKM has joined #openstack-trove08:02
*** flaper87|afk is now known as flaper8708:12
*** SergeyLukjanov has quit IRC08:46
*** yogesh_ has quit IRC09:52
*** haomaiwang has quit IRC09:52
*** decede has quit IRC09:52
*** konetzed has quit IRC09:52
*** yogesh_ has joined #openstack-trove09:54
*** haomaiwang has joined #openstack-trove09:54
*** decede has joined #openstack-trove09:54
*** konetzed has joined #openstack-trove09:54
*** rongze has quit IRC10:09
openstackgerritDenis M. proposed a change to openstack/trove: Block delete instance while backup  https://review.openstack.org/6196910:18
*** SergeyLukjanov has joined #openstack-trove10:24
*** rongze has joined #openstack-trove10:40
*** rongze has quit IRC10:45
openstackgerritDirk Mueller proposed a change to openstack/python-troveclient: Tighten flake8 checks in Gating  https://review.openstack.org/6214610:57
openstackgerritIonut Artarisi proposed a change to openstack/trove: don't try to concatenate a string with a variable that can be None  https://review.openstack.org/6233210:58
*** yogesh_ has quit IRC11:13
openstackgerritDenis M. proposed a change to openstack/trove: Block delete instance while backup  https://review.openstack.org/6196911:13
*** yogesh_ has joined #openstack-trove11:13
*** yogeshme_ has joined #openstack-trove11:17
*** yogesh_ has quit IRC11:18
*** yogeshme_ has quit IRC11:31
*** SushilKM has quit IRC11:34
*** SergeyLukjanov has quit IRC11:51
*** SergeyLukjanov has joined #openstack-trove11:52
openstackgerritDenis M. proposed a change to openstack/trove: Block delete instance while backup  https://review.openstack.org/6196912:09
*** rongze has joined #openstack-trove12:41
*** rongze has quit IRC12:46
*** nosnos has quit IRC12:47
*** dougshelley66 has joined #openstack-trove13:02
*** pdmars has joined #openstack-trove13:26
*** radez_g0n3 is now known as radez13:32
*** dougshelley66 has quit IRC13:45
*** zigo has quit IRC13:55
*** jasonb365 has joined #openstack-trove13:58
*** zigo has joined #openstack-trove14:00
*** rongze has joined #openstack-trove14:12
*** rustlebee is now known as russellb14:22
*** denis_makogon has quit IRC14:23
*** SergeyLukjanov is now known as _SergeyLukjanov14:24
*** _SergeyLukjanov has quit IRC14:24
*** vgnbkr has quit IRC14:24
*** jcru has joined #openstack-trove14:24
*** robertm__ has joined #openstack-trove14:26
*** robertmyers has joined #openstack-trove14:28
*** robertm__ has quit IRC14:28
*** robertmyers has quit IRC14:29
*** robertmyers has joined #openstack-trove14:30
*** dougshelley66 has joined #openstack-trove14:30
*** robertmyers has quit IRC14:30
*** denis_makogon has joined #openstack-trove14:30
*** robertmyers has joined #openstack-trove14:31
*** openstackgerrit has quit IRC14:31
*** robertmyers has quit IRC14:31
*** openstackgerrit has joined #openstack-trove14:31
*** robertmyers has joined #openstack-trove14:32
*** robertmyers has quit IRC14:32
*** robertmyers has joined #openstack-trove14:33
*** robertmy_ has joined #openstack-trove14:33
*** robertmyers has quit IRC14:33
*** robertmy_ has quit IRC14:34
*** robertmyers has joined #openstack-trove14:34
*** decede has quit IRC14:35
*** robertmyers has quit IRC14:35
*** robertmyers has joined #openstack-trove14:35
*** robertmyers has quit IRC14:36
*** robertmyers has joined #openstack-trove14:37
*** decede has joined #openstack-trove14:38
*** abramley has joined #openstack-trove14:43
*** rnirmal has joined #openstack-trove14:46
*** kevinconway has joined #openstack-trove14:53
*** kevinconway has quit IRC14:53
*** amytron has joined #openstack-trove14:57
*** jcooley_ has joined #openstack-trove15:04
*** glucas has joined #openstack-trove15:06
*** doug_shelley66 has joined #openstack-trove15:10
*** diakunchikov_ has joined #openstack-trove15:13
*** dougshelley66 has quit IRC15:14
*** diakunchikov has quit IRC15:14
*** demorris has joined #openstack-trove15:15
*** kpom_ has joined #openstack-trove15:16
*** SergeyLukjanov has joined #openstack-trove15:17
*** SushilKM has joined #openstack-trove15:18
*** vgnbkr has joined #openstack-trove15:24
*** kevinconway has joined #openstack-trove15:24
*** vipul has quit IRC15:27
*** kevinconway_ has joined #openstack-trove15:28
*** vipul has joined #openstack-trove15:28
*** kevinconway has quit IRC15:29
*** kevinconway_ is now known as kevinconway15:29
openstackgerritSushil Kumar proposed a change to openstack/trove: Replaces location to be backed up during prepare  https://review.openstack.org/6239915:30
*** greghill has joined #openstack-trove15:31
greghillso, does 'recheck no bug' not refire the reddwarf tests?  If not, how do I do that?15:31
SushilKMyou need to contact here in with someone who has got the access to re-run the reddwarf tests15:32
SushilKMthey have changed it now15:32
*** datsun180b has joined #openstack-trove15:35
greghillhttps://review.openstack.org/#/c/62039/ needs said reddwarf tests re-run15:35
*** zigo has quit IRC15:39
*** haomaiwang has quit IRC15:39
*** konetzed has quit IRC15:39
datsun180bgreghill: on it15:39
greghillgracias15:39
*** haomaiwang has joined #openstack-trove15:39
*** SushilKM has quit IRC15:43
*** zigo has joined #openstack-trove15:44
*** konetzed has joined #openstack-trove15:44
*** amytron_ has joined #openstack-trove15:45
*** jasonb365_ has joined #openstack-trove15:45
*** haomaiwa_ has joined #openstack-trove15:46
*** diakunchikov__ has joined #openstack-trove15:47
*** cp16net_ has joined #openstack-trove15:48
*** tanisdl has joined #openstack-trove15:50
*** datsun_F40PH has joined #openstack-trove15:50
*** SlickN1k has joined #openstack-trove15:51
*** jdbarry_ has joined #openstack-trove15:51
*** mtaylor has joined #openstack-trove15:52
*** pdmars_ has joined #openstack-trove15:52
*** plodronio has joined #openstack-trove15:52
*** dmakogon_ has joined #openstack-trove15:52
*** jasonb365 has quit IRC15:54
*** SlickNik has quit IRC15:54
*** rnirmal has quit IRC15:54
*** jdbarry has quit IRC15:54
*** datsun180b has quit IRC15:54
*** diakunchikov_ has quit IRC15:54
*** amytron has quit IRC15:54
*** mordred has quit IRC15:55
*** jdbarry_ is now known as jdbarry15:55
*** jasonb365_ is now known as jasonb36515:55
*** amytron_ is now known as amytron15:55
*** SlickN1k is now known as SlickNik15:55
*** datsun_F40PH is now known as datsun180b15:55
*** haomaiwang has quit IRC15:56
*** denis_makogon has quit IRC15:56
*** datsun180b has quit IRC15:57
*** datsun180b has joined #openstack-trove15:57
*** cp16net has quit IRC15:57
*** pdmars has quit IRC15:57
*** cp16net_ is now known as cp16net15:57
*** demorris_ has joined #openstack-trove16:01
*** jasonb365 has quit IRC16:01
*** demorris has quit IRC16:02
*** demorris_ is now known as demorris16:02
*** flaper87 is now known as flaper87|call16:02
*** datsun180b has quit IRC16:07
*** plodronio_ has joined #openstack-trove16:07
*** haomaiwang has joined #openstack-trove16:08
*** jcooley_ has quit IRC16:09
dmakogon_greghill, you also could ping me, i'm in different time zone UTC(+2), to restart reddwarf builder16:09
*** dmakogon_ is now known as denis_makogon16:09
*** jcooley_ has joined #openstack-trove16:11
kevinconwaydenis_makogon: the time change makes it really hard though. it takes longer for the bytes to travel.16:12
denis_makogonkevinconway, yup16:12
*** spiffxp_ has joined #openstack-trove16:14
*** plodronio has quit IRC16:18
*** mtaylor has quit IRC16:18
*** cp16net has quit IRC16:18
*** haomaiwa_ has quit IRC16:18
*** spiffxp has quit IRC16:18
*** jcooley has quit IRC16:18
*** demorris has quit IRC16:18
*** zigo has quit IRC16:18
*** konetzed has quit IRC16:18
*** hub_cap has quit IRC16:18
*** kiall has quit IRC16:18
*** annashen has quit IRC16:18
*** SlickNik has quit IRC16:18
*** SergeyLukjanov has quit IRC16:18
*** kpom_ has quit IRC16:18
*** greghill has quit IRC16:18
*** esmute has quit IRC16:18
*** metral has quit IRC16:18
*** plodronio_ is now known as plodronio16:18
*** cp16net has joined #openstack-trove16:21
*** demorris has joined #openstack-trove16:21
*** SlickNik has joined #openstack-trove16:21
*** konetzed has joined #openstack-trove16:21
*** zigo has joined #openstack-trove16:21
*** greghill has joined #openstack-trove16:21
*** kpom_ has joined #openstack-trove16:21
*** esmute has joined #openstack-trove16:21
*** hub_cap has joined #openstack-trove16:21
*** metral has joined #openstack-trove16:21
*** kiall has joined #openstack-trove16:21
*** jcooley has joined #openstack-trove16:21
*** annashen has joined #openstack-trove16:21
openstackgerritDenis M. proposed a change to openstack/trove: Show AZ in instance create response  https://review.openstack.org/6241416:21
openstackgerritDenis M. proposed a change to openstack/trove: Allow AZ at instance show  https://review.openstack.org/6241516:21
*** mtaylor has joined #openstack-trove16:25
*** rwsu_ has joined #openstack-trove16:25
*** rwsu has quit IRC16:27
*** kpom_ has quit IRC16:29
*** tanisdl has quit IRC16:39
*** tanisdl has joined #openstack-trove16:40
greghillhttps://review.openstack.org/#/c/62039/ needs some +2 love - fixes it so older instances don't break after you upgrade to datastore support16:49
*** SushilKM has joined #openstack-trove16:54
*** flaper87|call is now known as flaper8716:56
denis_makogonSushilKM, ping16:58
*** haomaiwang has quit IRC17:01
*** demorris has quit IRC17:03
SushilKMhey denis_makogon17:03
denis_makogonSushilKM, i already wrote several comments on your competitive review17:04
SushilKMyeah let me read the meeting comments and put my answer the way i think :)17:04
denis_makogonok17:06
*** demorris has joined #openstack-trove17:07
*** jcooley_ has quit IRC17:07
*** jasonb365 has joined #openstack-trove17:08
SushilKMvipul, slicknik, hub_cap plz have a look on this one https://review.openstack.org/#/c/62399/17:15
SushilKMany feedbacks welcome17:15
*** SnowDust has joined #openstack-trove17:17
denis_makogonSushilKM, since Trove would support more than one datastore type we need more flexible mechanism of configuration, that is why your suggestion is not acceptable it global terms, biggest issue - you suggest to create new N parameters for each datastore, imagine how guestagent.conf would look in this case17:19
denis_makogonwe need better way17:19
denis_makogonsomeday we discussed config template per datastore type17:20
denis_makogonSushilKM, https://blueprints.launchpad.net/trove/+spec/template-config-per-datastore17:20
SushilKMpoint taken denis_makogon17:21
denis_makogonand it's already taken by me17:21
SushilKMbut my bug addresses the current problem which is mysql prepare for trove17:21
denis_makogonconfig parameter renaming is not what we need17:22
SushilKMi would only suggest that blueprint implementation may follow and override any changes done previously, and we do not know when would come the blueprint implementation17:22
SushilKMso, until then instead of living the bug, we shud rather agree to resolve that asap17:22
denis_makogoni'm, working on implementing oslo conf groups, they would fix any kinds of issues with any configuration ever possible17:23
denis_makogonSushilKM, it's not a bug17:23
denis_makogonSushilKM, we call it smart configuration17:23
denis_makogonSushilKM, no estimates at community17:23
*** misteraa has joined #openstack-trove17:23
SushilKMhe he he "smart configuration" something new i learnt today ;)17:24
denis_makogonuntil new datastores wound not land it's not even a bug17:24
SushilKMits a bug17:24
SnowDustdenis makogon, lets bring more oslo17:24
SnowDusteven for rpc17:25
SnowDustfrom openstack.common << this should be cleaned up :)17:25
SushilKMplease read  description at https://bugs.launchpad.net/trove/+bug/1261354, i have dumped all the thoughts there.17:25
*** misteraa has quit IRC17:26
*** jmontemayor has joined #openstack-trove17:26
denis_makogonSushilKM, bug is incomplete, why does you ingore data dir in config.template ?17:28
SushilKMi m not ignoring config.template, it is only that "/var/lib/mysql" is the default location for the first installation of mysql, and thats the way it has been written in trove until now17:30
SushilKMok, I would explain here17:32
SushilKMIn between these lines https://github.com/openstack/trove/blob/master/trove/guestagent/datastore/mysql/manager.py#L119-L12117:32
SushilKMprepare method preserves data always from CONF.mount_point, which is not a valid case, if someone changes the CONF.mount_point value which obviously is not meant to point the data directory17:33
denis_makogonSushilKM, how do you plan to deal with each new datastore ?17:34
SushilKMjust a sec .... let me complete first @denis_makogon17:34
*** dlpe__ has joined #openstack-trove17:34
SushilKMThe data would always exist on /var/lib/mysql ir-respective of value of CONF.mount_point, because that is the location where mysql puts its data when it is installed the first time, using the commands specified in trove-integration image creation as well as install_if_needed method.17:35
denis_makogonSushilKM, why does mount point were renamed to mysql_dir ? When does it became mysql specific parameter ?17:35
SushilKMdid not get ur point @denis_makogon17:35
*** amytron_ has joined #openstack-trove17:36
denis_makogonwhy do you rename mount_point to mysql_dir ?17:36
SushilKMWe should be preserving /var/lib/mysql instead of CONF.mount_point.17:36
SushilKMok, its mysql_base_dir and mysql_dir, i got confused ....17:36
*** amytron has quit IRC17:37
*** amytron_ is now known as amytron17:37
denis_makogoni mean mysql_base_dir17:37
SushilKMand CONF.mount_point could be changed irrespective of data directory being renamed where we had initial data which needed to be backed up17:37
SushilKMas the code lines specified is used to rsync the data to the volume mounted17:38
denis_makogonwhat what kind of value-add brings your patch ?17:39
denis_makogonas for me it doesn't change anything17:39
SushilKMwhen user changes the value of CONF.mount_point, this patch would back the correct data directory to the mounted path17:40
SushilKMmounted volume rather17:40
denis_makogonuser cannot change mount point17:41
*** flaper87 is now known as flaper87|afk17:41
denis_makogonit has fixed value per deployment17:41
*** denis_makogon has quit IRC17:42
*** diakunchikov__ has quit IRC17:42
*** ikhudoshyn has quit IRC17:42
SushilKMi thought we can make a change in configuration file and update CONF.mount_point17:42
*** diakunchikov__ has joined #openstack-trove17:42
*** denis_makogon has joined #openstack-trove17:42
SushilKM i thought we can make a change in configuration file and update CONF.mount_point17:43
denis_makogonno, user cannot change configuration at all17:43
*** ikhudoshyn has joined #openstack-trove17:43
denis_makogononly administrator17:43
SushilKMby user i mean administrator only17:44
denis_makogonand each value is fixed per trove deployment17:44
SushilKMadmin is the user of trove code17:44
SushilKMso, the entire trove deployment could be the candidate17:45
denis_makogonso, you suggest to let users communicate with cloud provider administrators and ask them to change mount point, because the weather is not so good today ? =)17:45
SushilKMnot exactly17:45
denis_makogonSushilKM, production doesn't work that way17:45
SushilKMi m saying if being a cloud provider, for any other reason, i want my mount_point to be different, i have the option of using this parameter CONF.mount_point17:47
SushilKMbut when I use it, the biggest problem in the way is that i need to backup my data directory where mysql installed its database and that is obviously not the CONF.mount_point17:48
denis_makogonwhat's the point of migrating data from /var/lib/mysql to /var/lib/mysql ?17:49
SushilKMfrom where did question came17:50
denis_makogonyour suggestion bring rather big inconsistence: with your patch we would have 3 values: mount point, base_dir, data_dir17:50
denis_makogonSushilKM, but you didn't change anything17:50
denis_makogononly renaming17:50
SushilKMits not renaming17:51
SushilKMit is a new parameter which separates the data path which I named as mysql_base_dir(which points to location which hold mysql database's data)17:52
SushilKMthen I used that parameter to differentiate the path to be backed up while mounting a different mount_point17:52
denis_makogonSushilKM, data_dir hold mysql data17:53
SushilKMi wrote that in my patchse commit message17:53
*** harlowja_away is now known as harlowja17:53
denis_makogoni don't like "third leg"17:54
*** SergeyLukjanov has joined #openstack-trove17:54
SushilKMi did not saw that data_dir anywhere in the code17:54
*** jcooley_ has joined #openstack-trove17:54
denis_makogonSushilKM, https://github.com/openstack/trove/blob/master/trove/templates/mysql/config.template#L1117:57
*** demorris_ has joined #openstack-trove17:58
SushilKMbut thats in config templates, and the path we had data installed gets before this file is read, you can make changes in this file to point to new location of datafir, but you cannot change the original location of mysql data directory by making a change in this config template17:59
*** demorris has quit IRC18:00
*** demorris_ is now known as demorris18:00
denis_makogonyou can change it without ruining code18:00
denis_makogonjust change it in config.template18:00
*** SnowDust has quit IRC18:01
SushilKMbut this config.template gets in use when u inititate ur instance using the service image, but mysql was installed in ur serviceimage prior to it had any config.template18:02
denis_makogonwhy does mysql installed at service image ?18:04
SushilKMhttps://github.com/openstack/trove-integration/blob/master/scripts/files/elements/ubuntu-mysql/install.d/10-mysql says that18:07
denis_makogonSushilKM, do not look at integration at tall]18:07
denis_makogonSushilKM, int only for development and testing18:07
SushilKMthen hold a sec, i wud give u another one18:07
denis_makogonSushilKM, idea is to use pure cloud-init images18:08
denis_makogonnothing else18:08
SushilKMhttps://github.com/openstack/trove/blob/master/trove/guestagent/datastore/mysql/service.py#L592 installs mysql-server package and that gets us a database.18:09
*** amcrn has joined #openstack-trove18:09
*** SergeyLukjanov has quit IRC18:10
SushilKMor if this did not installed mysql then where did you get mysql from obviously from service-image-creation with pre-installled mysql18:11
SushilKMi can think of getting this mysql-server package only in two ways18:11
*** SergeyLukjanov has joined #openstack-trove18:12
denis_makogon1. cloud init 2. datastore version packages18:12
denis_makogononly that way18:12
robertmyersSushilKM: the guest should always install the datastore in production18:12
*** yidclare has joined #openstack-trove18:13
robertmyersthe image is not a good place to install that18:13
*** yogesh has joined #openstack-trove18:13
SushilKMfine that, guest would install but then it would have data dir at /var/lib/mysql irrespective of config.template18:13
robertmyersotherwise every security update will need a new image18:13
*** SnowDust has joined #openstack-trove18:15
robertmyersSushilKM: another point that if you install the datastore, then it *will* place the data in /var/lib/mysql so you would not be able to mount that in a volume18:15
robertmyersbecause the directory is not empty18:15
robertmyersso you effectively kill mysql18:16
robertmyersso be careful when messing with mount point18:16
SushilKMagree to the concept robertmyers18:17
SushilKMbut mounting volume problem i did not get that comment18:17
robertmyersyou can't mount a directory with data in it18:18
robertmyerstry it18:18
SushilKMi wud try that, get back with my finidings18:21
*** demorris has quit IRC18:21
denis_makogonrobertmyers, +1 for mounting18:21
*** demorris has joined #openstack-trove18:24
SushilKMwe cannot mount a directory with data, but that data would be rsyncd out, i think, though not tested18:26
denis_makogonrsync 'ed where ?18:27
SushilKMto tmp_mount_point18:27
*** iliyaw has joined #openstack-trove18:28
*** iliyaw has quit IRC18:28
*** SushilKM has quit IRC18:36
*** rongze has quit IRC18:41
*** rnirmal has joined #openstack-trove18:43
greghillso, is there a plan to separate the mysql-extension-specific tests into a separate folder or something at some point?18:48
*** liksey has joined #openstack-trove18:48
*** liksey has quit IRC18:49
openstackgerritDenis M. proposed a change to openstack/trove: Remove 'mount_point' from guestagent RPC API  https://review.openstack.org/5718919:00
openstackgerritSushil Kumar proposed a change to openstack/trove: Changes command to create the directory  https://review.openstack.org/6244819:01
denis_makogongreghill, which tests ?19:02
greghillsome of the api tests cover behavior specified by the mysql extension from what I can tell19:03
greghilljust wondered if we were gonna split that up to make it easier to test specific extensions19:03
*** glucas has quit IRC19:05
*** SushilKM has joined #openstack-trove19:05
SushilKMm back robertmyers19:06
SushilKMhave something19:07
openstackgerritDenis M. proposed a change to openstack/trove: Remove 'mount_point' from guestagent RPC API  https://review.openstack.org/5718919:07
SushilKMI used https://review.openstack.org/#/c/62399/ and https://review.openstack.org/#/c/62448/19:07
*** glucas has joined #openstack-trove19:08
SushilKMthen i followed this with a change in CONF.mount_point parameter to point it to somewhere else other than /var/lib/mysql, i took /var/lib/mysqlsrv19:08
SushilKMand the results are good19:08
SushilKMI got a volume mounted as /var/lib/mysqlsrv with the mysql database files inside that, instance getting active19:09
SushilKMand this is the guestagent log i got https://gist.github.com/sushilkm/799255619:10
denis_makogonSushilKM, but mysql itself writes its data to /var/lib/mysql19:10
denis_makogonSushilKM, try to backup it =)19:11
denis_makogonSushilKM, but first load some data into mysql19:11
denis_makogonanything19:11
SushilKMyeah it has the data files there19:14
SushilKMcoz, it has got its intial databases like performance_schema, mysql, etc.19:14
denis_makogonSushilKM, i suppose you launch guest from root19:16
*** plodronio has quit IRC19:16
denis_makogonSushilKM, do next steps: 1. Launch guest not from root user, 2. Load data into mysql,  3. Try to backup data  ----------------------------- do every step with your patchsets19:18
SushilKMlaunche guest from root, is that the way trove is code meant to be run?19:18
robertmyersSushilKM: no never19:19
SushilKM@denis_makogon, the step i have changed gets called during prepare method only19:19
SushilKMso, whats the use of backup restore manually19:19
denis_makogonSushilKM, not from root19:19
robertmyersdenis_makogon: xtrabackup and mysqldump use the my.cnf file to locate the data19:20
denis_makogonSushilKM, if you lauch guest from root, every operation would have UID = 0 (root), every task would be successful19:20
robertmyersnot mount point that we pass it19:20
robertmyersin xtrabackup that argument is ignored19:21
SushilKMhe he he, but trove is not written to be used with root i suppose19:21
robertmyersdenis_makogon: don't run the guest as root19:21
denis_makogonrobertmyers, point me19:21
SushilKMalso denis, the last command "if guest would be launched from simple user, this step would fail" i am addressing this itself as the bug19:21
denis_makogonrobertmyers, i never did it19:21
robertmyersits just a bad idea19:21
denis_makogonrobertmyers, i was telling SushilKM that he should not launch guest as root19:22
robertmyershe isn't19:22
*** jmontemayor has quit IRC19:22
*** glucas has quit IRC19:22
SushilKMrobertmyers, did u see my gist19:22
robertmyershe is adding a call to use sudo19:22
SushilKMi was able to change mount_point using my patchsets19:22
*** glucas has joined #openstack-trove19:23
*** SnowDust has quit IRC19:23
robertmyersSushilKM: good, I'm a little lost with all these changes19:24
denis_makogonSushilKM, ok, i suggest to make dependent reviews19:24
SushilKMjust to breif i made two changes19:24
robertmyersI'm not sure what we are trying to do anymore :)19:24
denis_makogonit would be better to track them19:24
robertmyerssplit mount_dir from data_dir?19:24
SushilKMif we change CONF.mount_point to get it mounted to a different location, we get a problem at https://github.com/openstack/trove/blob/master/trove/guestagent/datastore/mysql/manager.py#L119-L12119:25
SushilKMbecause it backs up CONF.mount_point19:25
denis_makogonSushilKM, https://review.openstack.org/#/c/57189/19:26
robertmyersSushilKM: only if you install mysql on the image19:26
SushilKMso i changed it with a different parameter CONF.mysql_base_dir(if there are issues with name, i may change that)19:26
SushilKMthen next i had was i was not getting directory created because of os.makedirs which i replaced with sudo19:27
SushilKMadded both these changes to my trove followed by a change to CONF.mount_point point something other than /var/lib/mysq19:27
denis_makogonSushilKM, i'm ok, with sudo, but disagree with adding another one conf parameter19:28
SushilKMand then the results i have populated in gist19:28
*** jmontemayor has joined #openstack-trove19:28
denis_makogonrobertmyers, https://review.openstack.org/#/c/57189/1819:28
denis_makogonSushilKM, have you read my patchset  scope ?19:29
SushilKMyeah i saw that, thats broad, and different19:31
robertmyersdenis_makogon: this looks good, however, it seems that this really isn't mount_point, but rather data_dir19:33
robertmyersyou only mount if you use volumes19:33
robertmyersso, I think we need to think about what we want to do here19:34
denis_makogoni used same parameter19:34
robertmyersI see the new cfg group19:34
denis_makogonrobertmyers, yes, for mysql19:35
robertmyersit seems that we have abused the mount_point in the code in place of data_dir19:35
robertmyerswe just have gotten by without problems as they are set to the same value19:36
robertmyersbut in reality they *could * be different19:36
denis_makogonrobertmyers, yes, it would be next patchset19:39
*** konetzed has quit IRC19:43
*** konetzed has joined #openstack-trove19:44
SushilKMtests are failing for https://review.openstack.org/#/c/62448/ need to make the change there too, tests are for os.makedirs :(19:45
*** rnirmal has quit IRC19:46
openstackgerritSushil Kumar proposed a change to openstack/trove: Changes command to create the directory  https://review.openstack.org/6244819:50
SushilKMhey robertmyers i have edited the tests on that checkin :)19:52
SushilKMhey hub_Cap, slicknik, vipul, robertmyers needed reviews on few checkins of mine :) https://review.openstack.org/#/c/61948/     https://review.openstack.org/#/c/62050/  https://review.openstack.org/#/c/62235/  https://review.openstack.org/#/c/62236/  https://review.openstack.org/#/c/62399/  https://review.openstack.org/#/c/62448/19:58
*** denis_makogon has quit IRC19:58
SushilKMmany thanks for ur helping feedbacks and reviews :D19:58
*** denis_makogon has joined #openstack-trove19:58
*** demorris has quit IRC20:01
*** SushilKM has quit IRC20:02
*** demorris has joined #openstack-trove20:03
*** plodronio has joined #openstack-trove20:22
*** jcooley_ has quit IRC20:28
*** freyes has joined #openstack-trove20:36
*** amrith has joined #openstack-trove20:53
*** jcooley_ has joined #openstack-trove20:56
openstackgerritFelipe Reyes proposed a change to openstack/trove: Adds missing dependency to the documentation  https://review.openstack.org/6013320:58
*** amrith has quit IRC20:59
*** radez is now known as radez_g0n321:04
*** freyes has quit IRC21:05
*** demorris_ has joined #openstack-trove21:06
*** demorris has quit IRC21:07
*** demorris_ is now known as demorris21:07
*** openstackgerrit has quit IRC21:08
*** openstackgerrit has joined #openstack-trove21:08
*** denis_makogon_ has joined #openstack-trove21:09
*** denis_makogon has quit IRC21:09
*** denis_makogon_ is now known as denis_makogon21:09
*** dmakogon_ has joined #openstack-trove21:10
denis_makogonwhazzaap 2 all21:10
*** jcooley_ has quit IRC21:13
*** jcooley_ has joined #openstack-trove21:19
*** abramley has quit IRC21:19
*** radez_g0n3 is now known as radez21:20
*** glucas has quit IRC21:20
*** jasonb365 has quit IRC21:29
*** jcooley_ has quit IRC21:32
*** jasonb365 has joined #openstack-trove21:33
*** demorris has quit IRC21:33
*** konetzed has quit IRC21:34
*** konetzed has joined #openstack-trove21:34
*** demorris has joined #openstack-trove21:35
*** SergeyLukjanov has quit IRC21:35
*** SergeyLukjanov has joined #openstack-trove21:38
*** SergeyLukjanov_ has joined #openstack-trove21:42
*** SergeyLukjanov is now known as SergeyLukjanov_a21:42
*** SergeyLukjanov_a has quit IRC21:43
*** jasonb365 has quit IRC21:45
*** abramley has joined #openstack-trove21:47
*** radez is now known as radez_g0n321:48
*** jasonb365 has joined #openstack-trove21:49
*** SergeyLukjanov_ has quit IRC21:57
*** SergeyLukjanov has joined #openstack-trove21:57
*** kevinconway has quit IRC21:59
*** jcooley_ has joined #openstack-trove22:00
*** SergeyLukjanov has quit IRC22:04
*** pdmars_ has quit IRC22:09
*** doug_shelley66 has quit IRC22:28
denis_makogonSlickNik, could you please re-review this one ? https://review.openstack.org/#/c/58845/22:32
SlickNikdenis_makogon: sure, will take a look. Thanks!22:32
denis_makogonSlickNik, would be nice to get this one reviewed too https://review.openstack.org/#/c/61969/22:34
denis_makogonSlickNik, as you remember, we spoke about mount_point and data_dir22:35
denis_makogonSlickNik, i made few steps to make this situation a bit clear https://review.openstack.org/#/c/57189/22:36
*** rnirmal has joined #openstack-trove22:36
SlickNikSeems like Jenkins doesn't like your changes on that one, you might want to appease him first.22:37
*** rnirmal has quit IRC22:38
denis_makogonSlickNik, it would change whole code, only tests22:39
denis_makogonSlickNik, i'd like to hear thoughts about main idea of patchset22:39
SlickNikSo, one thing I notice is that the patch is still breaking the guestagent RPC API.22:41
SlickNikI know vipul had some concerns regarding that as well.22:41
SlickNikIt's something that we're trying not to do (unless _really_ necessary)22:42
denis_makogonSlickNik, unfortunately, there's no way to leave guest API as it is22:42
denis_makogonSlickNik, because problem goes through taskmanager (which is hard to make general) and lands at guest22:43
denis_makogonalso, unfortunately, alot patchsets are making taskmanager code unsuitable for other datastores22:44
SlickNikvipul is out today. Let's talk about it again tomorrow when vipul is around. I'd like to get his input on it as well.22:44
denis_makogonok22:44
SlickNikdenis_makogon: can you clarify what you mean by "alot patchsets are making taskmanager code unsuitable for other datastores"?22:45
SlickNik(perhaps with an example?)22:45
denis_makogonexample: https://review.openstack.org/#/c/59410/22:46
denis_makogonas for me, taskmanager should stay generic as it possible22:47
denis_makogonbut, unfortunately, for some reasons, we still ignoring this requirement22:47
denis_makogonTrove would support much more than it does now, we should deal with it22:48
denis_makogonand do not let patchsets ruine idea of generic server side and custom VM side22:50
*** flaper87|afk is now known as flaper8722:50
denis_makogonAnd personaly, i like to keep everything mysql-specific in guest code, otherwise trove would be mysql-provisioning-tool22:55
denis_makogonand that's sad =(22:56
*** robertmyers has quit IRC22:56
*** konetzed has quit IRC22:58
denis_makogonSlickNik, about AZ, trove back-end stores security groups/rules, nova/neutron does the same22:58
denis_makogonSlickNik, so why do not let trove back-end store single column for instance model ?22:59
*** konetzed has joined #openstack-trove23:00
*** dougshelley66 has joined #openstack-trove23:00
SlickNikdenis_makogon: having made a mistake in the past doesn't justify making a similar mistake in the future.23:01
SlickNikdenis_makogon: trove does not store instance IPs for example, why not follow that example instead?23:01
SlickNikIf there's a reason not to, I'd be open to discussing it. But let's make an effort to at least have a clean design.23:02
*** rnirmal has joined #openstack-trove23:02
*** jcooley_ has quit IRC23:04
*** denis_makogon has quit IRC23:04
*** jcooley_ has joined #openstack-trove23:06
cweidSo whats with the new H701  Empty localization string error in tox tests..23:06
cweidHow do I fix that error?23:06
cweidthe string is not empty..23:07
cweidI probably just don't understand the error...23:07
*** demorris has quit IRC23:07
SlickNikcweid: not sure. Do you have a log for me?23:12
SlickNikcweid: This might help: https://github.com/openstack-dev/hacking/blob/master/hacking/core.py#L75623:12
*** jcooley_ has quit IRC23:14
*** greghill has quit IRC23:16
cweidSlickNik: what I get from tox https://gist.github.com/cweidenkeller/799655323:16
cweidThe error is just not very descriptive...23:16
cweidthe _() local stuff only takes one arg right?23:16
SlickNikcweid: is there a stack trace before that error? Can you paste the whole tox log in a gist?23:17
cweidSlickNik: Reload that same gist.23:18
SlickNikwow, that's not very helpful23:19
cweidSlickNik: found some docs on it...23:22
cweidhttp://docs.openstack.org/developer/hacking/23:22
cweidsorry this is better http://docs.openstack.org/developer/hacking/#internationalization-i18n-strings23:23
*** jcooley_ has joined #openstack-trove23:26
openstackgerritConrad Weidenkeller proposed a change to openstack/trove: Added Redis Crud Operations blueprint redis-support  https://review.openstack.org/5348123:27
*** plodronio has quit IRC23:30
*** jasonb365 has quit IRC23:30
*** jcooley_ has quit IRC23:30
*** zacksh_ is now known as zacksh23:33
*** jcooley_ has joined #openstack-trove23:39
*** jcooley_ has quit IRC23:42
*** plodronio has joined #openstack-trove23:46
*** jcooley_ has joined #openstack-trove23:47
*** yogeshmehra has joined #openstack-trove23:47
*** flaper87 is now known as flaper87|afk23:58
*** jmontemayor has quit IRC23:58
*** yogeshmehra has quit IRC23:59
« Sunday, 2013-12-15 Index Tuesday, 2013-12-17 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!