Tuesday, 2016-01-05

*** haomaiwang has quit IRC00:01
*** jengeb has quit IRC00:01
*** haomaiwang has joined #openstack-trove00:01
*** ashleighfarnham has quit IRC00:23
*** sabeen3 has quit IRC00:30
*** tosky has quit IRC00:37
*** cheneydc has joined #openstack-trove01:00
*** haomaiwang has quit IRC01:01
*** haomaiwang has joined #openstack-trove01:01
*** sonali5 has joined #openstack-trove01:28
*** imandhan has quit IRC01:28
*** pmackinn has quit IRC01:37
*** haomaiwang has quit IRC02:01
*** haomaiwang has joined #openstack-trove02:01
*** haomaiwang has quit IRC03:01
*** haomaiwang has joined #openstack-trove03:01
openstackgerritMerged openstack/python-troveclient: Remove py26 support  https://review.openstack.org/25247003:03
openstackgerritAlex Tomic proposed openstack/trove-specs: Vertica cluster grow and shrink  https://review.openstack.org/26353903:06
*** links has joined #openstack-trove03:52
*** haomaiwang has quit IRC04:01
*** haomaiwang has joined #openstack-trove04:01
*** imandhan has joined #openstack-trove04:12
*** haomaiwang has quit IRC04:13
*** imandhan has quit IRC04:17
*** imandhan has joined #openstack-trove04:18
*** imandhan has quit IRC04:22
*** ThunderEmperor has joined #openstack-trove04:22
*** cheneydc has quit IRC04:46
*** haomaiwang has joined #openstack-trove04:53
*** HimanshuGarg has joined #openstack-trove04:53
*** HimanshuGarg has quit IRC04:56
*** haomaiwang has quit IRC05:01
*** haomaiwang has joined #openstack-trove05:01
*** ThunderEmperor_ has joined #openstack-trove05:16
*** ThunderEmperor has quit IRC05:16
*** ThunderEmperor_ is now known as ThunderEmperor05:16
*** cheneydc has joined #openstack-trove05:23
openstackgerritMerged openstack/trove: Fix missing value types for log message  https://review.openstack.org/22300005:25
*** sgotliv has joined #openstack-trove05:39
*** haomaiwang has quit IRC06:01
*** haomaiwang has joined #openstack-trove06:01
*** coolsvap|away is now known as coolsvap06:02
*** coolsvap is now known as coolsvap|away06:17
*** coolsvap|away is now known as coolsvap06:20
*** haomaiwang has quit IRC07:01
*** 18WABPMMU has joined #openstack-trove07:01
*** ThunderEmperor has quit IRC07:07
openstackgerritShu Muto proposed openstack/python-troveclient: Drop py33 support  https://review.openstack.org/25771307:17
sonali5hi anybody there need a help07:20
sonali5?07:20
*** coolsvap is now known as coolsvap|away08:00
*** 18WABPMMU has quit IRC08:01
*** haomaiwang has joined #openstack-trove08:01
*** sgotliv has quit IRC08:06
*** sonali5 has quit IRC08:15
openstackgerrityapeng Yang proposed openstack/trove: Fix grammatical mistake, Changed character from "an" to "a"  https://review.openstack.org/26362808:27
*** tobe has joined #openstack-trove08:41
*** tobe has quit IRC08:41
*** haomaiwang has quit IRC08:53
*** haomaiwang has joined #openstack-trove08:53
*** haomaiwang has quit IRC09:01
*** haomaiwang has joined #openstack-trove09:01
*** cheneydc has quit IRC09:17
*** coolsvap|away is now known as coolsvap09:17
*** cheneydc has joined #openstack-trove09:18
*** sonali5 has joined #openstack-trove09:48
openstackgerritChaozhe Chen proposed openstack/trove: Trivial: replace deprecated _impl_messaging  https://review.openstack.org/26365509:58
*** cheneydc has quit IRC10:00
*** haomaiwang has quit IRC10:01
*** haomaiwang has joined #openstack-trove10:01
*** sonali5 has quit IRC10:17
*** tosky has joined #openstack-trove10:22
*** ThunderEmperor has joined #openstack-trove10:26
*** jengeb has joined #openstack-trove10:27
openstackgerritOpenStack Proposal Bot proposed openstack/python-troveclient: Updated from global requirements  https://review.openstack.org/25653010:30
*** haomaiwang has quit IRC11:01
*** haomaiwang has joined #openstack-trove11:01
*** sgotliv has joined #openstack-trove11:13
*** tellesnobrega_af is now known as tellesnobrega11:43
*** coolsvap is now known as coolsvap|away11:54
*** haomaiwang has quit IRC12:01
*** haomaiwang has joined #openstack-trove12:01
*** cheneydc has joined #openstack-trove12:47
*** haomaiwang has quit IRC13:01
*** haomaiwang has joined #openstack-trove13:01
*** pmackinn has joined #openstack-trove13:08
*** links has quit IRC13:10
*** haomaiwang has quit IRC13:12
*** _amrith_ is now known as amrith13:24
amrithsushilkm_, my question was this. if you like, I can work on the root-disable thing.13:32
openstackgerritamrith proposed openstack/trove: Send up guest agent logs over the conductor channel to trove controller  https://review.openstack.org/23771013:33
*** haomaiwang has joined #openstack-trove13:40
*** sonali5 has joined #openstack-trove13:41
*** ThunderEmperor has quit IRC13:42
*** ThunderEmperor has joined #openstack-trove13:42
*** haomaiwang has quit IRC13:45
*** 18VAALI1N has joined #openstack-trove14:07
*** pmackinn has quit IRC14:23
*** links has joined #openstack-trove14:24
*** sonali5 has quit IRC14:30
*** kbyrne has quit IRC14:33
*** sonali5 has joined #openstack-trove14:35
*** jengeb has quit IRC14:37
*** jengeb1 has joined #openstack-trove14:37
*** kbyrne has joined #openstack-trove14:40
*** links has quit IRC14:49
*** kbyrne has quit IRC14:52
*** kbyrne has joined #openstack-trove14:53
*** kbyrne has quit IRC14:55
*** sbadia has left #openstack-trove14:58
*** kbyrne has joined #openstack-trove15:00
*** 18VAALI1N has quit IRC15:01
*** kbyrne has quit IRC15:01
*** haomaiwang has joined #openstack-trove15:01
*** kbyrne has joined #openstack-trove15:02
*** cweid has joined #openstack-trove15:03
*** pmackinn has joined #openstack-trove15:17
*** cheneydc has quit IRC15:27
*** sonali5 has quit IRC15:29
*** sonali5 has joined #openstack-trove15:39
*** csoukup has joined #openstack-trove15:49
openstackgerritDuk Loi proposed openstack/trove-dashboard: Add missing .gitreview file  https://review.openstack.org/26378015:50
*** sonali5 has quit IRC15:59
*** haomaiwang has quit IRC16:01
*** haomaiwang has joined #openstack-trove16:01
amrithpmackinn, g'morning16:08
amrithwhen you get a chance, would like to chat re: https://review.openstack.org/#/c/256079/16:08
pmackinnamrith, g'day16:12
pmackinnamrith, shirley16:12
amrithpmackinn, temple16:12
amrithbut I thought that was our safe-word.16:12
amrithwhat's shaking?16:13
*** shakamunyi has joined #openstack-trove16:14
*** barra204 has quit IRC16:15
pmackinnamrith, my -1? figured it might cost me a future +116:17
amrithwell, I was thinking a finger or three.16:17
pmackinnlol16:17
amrithGuido was dispatched this morning16:17
pmackinnhey now16:17
amrithhe works for the waste management company16:17
amrithwatch out for the dump truck as it backs up towards you.16:18
pmackinnmy -1 are powerless iirc16:18
amrithso what's three fingers?16:18
amrithone good drink?16:18
amrithno, seriously though16:18
amrithyour -1 is NOT powerless16:19
amriththe question is this ...16:19
amrithor the questions are these16:20
amrith1. there was no mail to the ML, I could send one.16:20
amrith2. I'm not sure I understand the second question16:20
amrithhbase has no SQL interface (that may be Hive)16:20
amrithhbase offers rest and thrift16:20
amrithyou could layer pig and hive on it16:20
amrithif you want16:20
amriththe issue I'm addressing is that some have asked us for support for hbase16:21
amrithand it makes sense, like adding support for a graph database (in the works now).16:21
pmackinncrap, got my hive and hbase mixed up...apologies16:21
pmackinnbut sahara is definitely in the business of hadoop clusters16:22
amriththat's ok. pig is the other option and I am wondering how to politely make a porcine reference but can't come up with one.16:23
amrithyes, sahara is in the business of hadoop clusters.16:23
amrithno doubt.16:23
amrithhbase is hadoop clusters plus a lot more16:23
pmackinnwhat are your feelings about api v1 fit for what hdfs represnets as a db?16:23
amrithso surely the argument could be made that trove should call sahara16:23
amrithbut that seems largely pointless given that much of the framework we need is already in Trove.16:23
amrithre: hdfs as a DB, I'm not really seeing HDFS as a DB.16:24
amrithHDFS is a storage subsystem as far as I can tell.16:24
amrithI can't 'query' hdfs any more than I can query ext3 (or 4)16:24
amrithmy understanding of HDFS is that it is a scalable distributed resilient <insert buzzword here> FILE SYSTEM.16:25
pmackinnbut you don't really resize it either, or back it up16:25
amrithsure you do resize HDFS16:25
amrithand sure you make data resilient16:25
amrithin a different way.16:25
amrithbut consider this, cassandra support the same n-copies mechanism.16:25
*** ThunderEmperor has quit IRC16:26
amrithso would you say that Trove shouldn't support backup for cassandra?16:26
amrithbackup serves a different function (IMHO) than n-copies and resiliency.16:26
amrithin point of fact, some may use replication as a backup and restore strategy16:26
pmackinnyou add/remove clusters and change replication16:26
amrithbut 'backup' in trove serves another purpose as well, cloning an instance.16:26
amrithcorrect, and in hbase fully-distributed, I intend that the cluster API will provide that capability as well.16:27
amrithbut for HBASE16:27
amrithor rather, with theprimary target being HBASE.16:27
amrithunder the covers, it means HDFS fully distributed as well16:27
pmackinnsure16:27
amrithbut that's only in service of HBase16:27
amrithIf you want a HDFS file system, then Trove isn't your answer.16:27
amrithIf you want HBase, I would like Trove to be a feasible solution16:27
pmackinnwell you mention this as a beachhead for other BD db16:28
amrithjust as it is if you want a cassandra cluster.16:28
amrithwhat is BD?16:28
pmackinnbig data16:28
amrithsure16:28
amrithvertica is a BD database16:28
amrithwe already support that.16:28
amrithand in any event, we do have people interested in another hadoop based database in Trove.16:29
amrithso yes, I see support for those as a natural extension for Trove.16:30
amrithinsofar as they are databases.16:30
amrithHDFS is NOT a database16:30
amritha Hadoop Cluster is not IMHO a database16:30
amrithif someone else wishes to add native hadoop cluster support to Trove, I wouldn't say no for sure.16:30
amrithif there is a use-case and someone would like to contribute code, why would I object to it?16:30
pmackinnwhich hbase flavor do you intend to support?16:31
pmackinnjust apache?16:32
amrithFor now, Apache HBase 1.1.216:32
amriththat's current16:32
*** ThunderEmperor has joined #openstack-trove16:32
pmackinnwhat about other distros? cloudera? hortonworks?16:33
pmackinnthese are roads already well travelled by sahara16:33
toskybut wouldn't it make sense to leave the provisioning part to Sahara, get the endpoints returned by Sahara and then let Trove do the things it's good for (operate with the database)?16:33
pmackinnamrith, also why make note of the different hbase clients other than their default ports?16:34
pmackinnstill command&control only, right?16:34
amrithpmackinn, for now, nothing. but great idea, maybe those too down the road.16:35
amrithsure tosky, that's an alternative.16:35
amrithnot one I prefer because it makes more stuff go out of Trove's control. for example, how would you implement a trove guest agent for that? would your image have two guest agents?16:36
toskySahara has no guest agent16:36
amrithpmackinn, what do you mean by difereht clients and ports.16:36
amrithtosky, ok. if sahara has no guest agent then where would you inject the trove guest agent?16:37
toskycouldn't it be a separate node in the cluster, oracle-like?16:37
amrithsure, all alternatives.16:37
pmackinn HBase Shell Support         |     Y      |      X     |      X     |16:38
pmackinn+-----------------------------+------------+------------+------------+16:38
pmackinn| REST API Support            |     Y      |      X     |      X     |16:38
pmackinn+-----------------------------+------------+------------+------------+16:38
pmackinn| Thrift API Support          |     Y      |      X     |      X     |16:38
pmackinn+-----------------------------+------------+------------+------------+16:38
pmackinn| Python Trove Client (CLI)   |     Y      |      X     |      X     |16:38
pmackinnwhat do these matter beside ports in the sec group? ^^16:38
amrithpmackinn, in order for hbase to support these, the daemons must be started16:38
amrithsee the trove-integration chkin which launches rest and thrift daemons.16:38
amrithand configures the ports in the cnf files16:38
toskythe "out of control" part is relative, as Sahara is another big-tent project, which is not going away, with a stable API16:38
amrithand as you say, puts them in the sec group.16:39
amrithsure tosky, nothing you say is wrong. it is an alternative.16:39
amriththis too, is an alternative approach.16:39
amrithand being a big-tent project makes no guarantees on longevity.16:40
amrithpmackinn, does that answer your question? re: support for clients.16:41
*** tosky has quit IRC16:42
pmackinni guess for me fundamentally: trove gets compute from nova, volume storage from cinder, swift storage for backups...seems sensible to interface with sahara for hadoop infra16:42
*** tosky has joined #openstack-trove16:42
*** ashleighfarnham has joined #openstack-trove16:43
pmackinnalso sahara is adding (added?) support for in-memory analytics like spark16:43
amrithI'm not sure that it seems any more 'sensible' than the argument that Trove should interface with heat for everything.16:43
amrithgreat, let sahara do that16:43
amrithmore options is good, or so I thought.16:43
amrithlike adding support for ceph and swift.16:44
toskyamrith: sorry, but the "no guarantees on longevity" is not an argument, or then it could open door to reimplementing every other service16:44
amrithit isn't an argument, I was just pointing out an inconsistency in your statement.16:44
toskyI'm pretty sure that if Sahara would die, that would mean that HBase support (and any other bigdata stuff) would not be needed anymore16:44
toskyI disagree on the inconsistency; code duplication on the other side is bad, and this is what the main proposal is leading to16:45
toskybut, anyway16:45
amrithtosky, feel free to put comments on the review however please be clear to observe that the things you suggest are alternative approaches which while valid do not detract from the validity of a different approach as proposed in the spec.16:48
toskyamrith: which means that it's pointless to add them16:49
amrithtosky, that's a defeatist point of view that I don't subscribe to. I vote (even though I get only one vote in 300+ million), I contribute to charity even though I'm not a millionaire, and I know people who exercise reguarly knowing that they have many generations of heart disease in their families. I also put -1's on reviews and sometimes +1's.16:50
amrithbut if you wish not to post comments, that is your choice.16:51
amrithBut, I believe the robustness of open source depends on fee and open communicatin and the mechanism of reviews. And therefore, while I disagree with your point of view, I respect it and welcome it as I do Pete's review.16:52
amrithpmackinn, what's the API fitness question?16:55
pmackinnamrith, e.g., what does it mean to backup Hbase?16:58
amrithit means, in Trove-speak, to serialize all the data in a form that can be used to launch a new instance at a future time.16:59
amrithand to store that on some inexpensive medium16:59
amrithsuch as swift16:59
amrithwhich is traditionally cheaper to operate than, say, cinder.16:59
amrithand the ability to do that in such a way that many such copies of a single instance can be made over time.17:00
amrithand in the future, extend that capability to fully-distributed hbase deployments with trove.17:00
*** haomaiwang has quit IRC17:01
amrithaka, same thing as it means for MySQL, PostgreSQL, MariaDB, MongoDB, Cassandra, CouchDB, Couchbase, Redis, ...17:01
*** haomaiwang has joined #openstack-trove17:01
amrithWhat is implemented (now) in the strategy is what is called Full Shutdown Backup in HBase terminology. See more at http://hbase.apache.org/0.94/book/ops.backup.html17:06
openstackgerritamrith proposed openstack/trove-specs: Add support for hbase in Trove  https://review.openstack.org/25607917:11
amrithpmackinn, I've updated the review with a section describing this backup and placing the link (above) in the references section.17:11
pmackinnamrith, thx17:11
sushilkm_@amrith go ahead, if you want to work on root-disable17:13
dloiHello @cpnet16.  Do you have the power to add the new trove-dashboard to launchpad?  Or do you want to open bugs and stuff in trove for that?17:13
amrithsushilkm_, thx. I will (most likely) ping you for some clarifications on what you had done in the bp you said may have been written already and to get more details about the bug.17:13
pmackinnamrith, i think it's the API fitness to the fully distributed model where things break down for me; "instance" is problematic since you are not really restoring an "instance"17:13
amrithpmackinn, so what of a pxc cluster?17:14
amrithlet's leave hbase out of this for a second and talk pxc cluster17:14
amrithlet's say mariadb with galera17:14
amrithwhat does it mean to take a backup of that?17:14
dloioops @cp16net I meant in the last entry for me :)17:14
amrithpmackinn, I thought the clarification re: instance and cluster in what I put in https://review.openstack.org/#/c/256079/7..8/specs/mitaka/hbase-in-trove.rst would address the issue you raise. so I must be missing something.17:16
amrithpmackinn, to be continued. I'm going to go and get something to eat before a 1pm meeting.17:17
pmackinnamrith, sure thx17:18
*** imandhan has joined #openstack-trove17:42
*** ThunderEmperor has quit IRC17:59
*** haomaiwang has quit IRC18:01
*** haomaiwang has joined #openstack-trove18:01
*** sabeen1 has joined #openstack-trove18:04
cp16netdloi: i'lll look18:14
*** sgotliv has quit IRC18:19
*** johnma has joined #openstack-trove18:20
cp16netdloi: i was looking around to see what other projects have done reguarding their dashbaoard and it looks like most of them have just filed the bugs/specs within their main lp project (designate and sahara)18:37
cp16netthey tag the bugs with dashboard to separate them in the list18:38
cp16netand put something in the bug/bp title as well18:38
cp16netlike [ui] - bug name18:38
cp16netor designate-dashboard: bug name18:38
cp16neti'll bring this up tomorrow in our meeting18:39
*** sabeen1 has quit IRC18:46
*** tosky has quit IRC18:48
*** tosky has joined #openstack-trove18:50
openstackgerritIshita Mandhan proposed openstack/trove: Code refactoring for couchdb  https://review.openstack.org/21017018:54
openstackgerritMerged openstack/trove: Define 'device_path' for Redis  https://review.openstack.org/26232418:57
*** haomaiwang has quit IRC19:01
*** haomaiwang has joined #openstack-trove19:01
*** jdbarry_ has joined #openstack-trove19:06
openstackgerritMerged openstack/trove: Replace assertEqual(None, *) with assertIsNone in tests  https://review.openstack.org/25903319:11
openstackgerritMerged openstack/python-troveclient: Updated from global requirements  https://review.openstack.org/25653019:16
*** davideagnello has quit IRC19:21
*** tellesnobrega is now known as tellesnobrega_af19:22
openstackgerritMayuri Ganguly proposed openstack/trove: WIP:Update Management Console user and password  https://review.openstack.org/20867019:42
*** sabeen1 has joined #openstack-trove19:46
*** davideagnello has joined #openstack-trove19:47
*** sabeen3 has joined #openstack-trove19:47
*** sabeen1 has quit IRC19:50
*** ShayneBurgess has joined #openstack-trove19:58
*** haomaiwang has quit IRC20:01
*** haomaiwang has joined #openstack-trove20:01
*** ShayneBurgess_ has joined #openstack-trove20:02
*** shayneburgess1 has joined #openstack-trove20:04
*** ShayneBurgess has quit IRC20:05
*** ShayneBurgess_ has quit IRC20:21
*** _mattf is now known as mattf20:43
*** mattf has quit IRC20:43
*** mattf has joined #openstack-trove20:43
*** mattf is now known as _mattf20:43
openstackgerritSimon Chang proposed openstack/trove: Fix to_gb & to_mb conversion function rounding issue  https://review.openstack.org/26390520:59
*** haomaiwang has quit IRC21:01
*** haomaiwang has joined #openstack-trove21:01
*** shayneburgess1 has quit IRC21:06
*** shayneburgess has joined #openstack-trove21:07
amrithpmackinn, hello21:07
*** chlong has quit IRC21:22
*** tellesnobrega_af is now known as tellesnobrega21:41
openstackgerritIshita Mandhan proposed openstack/trove: Code refactoring for couchdb  https://review.openstack.org/26392221:46
*** shayneburgess has quit IRC21:56
*** shayneburgess has joined #openstack-trove21:56
*** shayneburgess1 has joined #openstack-trove21:59
*** shayneburgess has quit IRC21:59
openstackgerritIshita Mandhan proposed openstack/trove: Code refactoring for couchdb  https://review.openstack.org/21017022:00
*** sgotliv has joined #openstack-trove22:00
*** haomaiwang has quit IRC22:01
*** haomaiwang has joined #openstack-trove22:01
*** shayneburgess has joined #openstack-trove22:01
*** shayneburgess1 has quit IRC22:01
*** shayneburgess has quit IRC22:04
*** shayneburgess has joined #openstack-trove22:04
*** jdbarry_ has quit IRC22:04
*** shayneburgess1 has joined #openstack-trove22:06
*** shayneburgess has quit IRC22:06
*** shayneburgess has joined #openstack-trove22:16
*** shayneburgess1 has quit IRC22:18
*** shayneburgess1 has joined #openstack-trove22:18
*** shayneburgess has quit IRC22:18
cp16netpmalik: so looks like the cluster scenario tests are working for pxc after i rebased my pxc grow/shrink on your finish cluster int-tests patch22:21
cp16neti just had to remove the skips for grow shrink on pxc22:22
*** haomaiwang has quit IRC22:23
openstackgerritSimon Chang proposed openstack/trove: Fix to_gb & to_mb conversion function rounding issue  https://review.openstack.org/26390522:26
*** haomaiwang has joined #openstack-trove22:28
*** sushilkm_ has quit IRC22:43
*** shayneburgess1 has quit IRC22:45
pmalikcp16net good news. I was actually testing with pxc as well (not grow/shrink at the time).22:49
*** haomaiwang has quit IRC23:01
*** haomaiwang has joined #openstack-trove23:01
*** chlong has joined #openstack-trove23:12
*** csoukup has quit IRC23:12
*** chlong has quit IRC23:14
*** tellesnobrega is now known as tellesnobrega_af23:27
*** georgelorch has quit IRC23:28
*** chlong has joined #openstack-trove23:30
*** sgotliv has quit IRC23:32
*** georgelorch has joined #openstack-trove23:42
*** sabeen3 has quit IRC23:45
openstackgerritMerged openstack/python-troveclient: Drop py33 support  https://review.openstack.org/25771323:47
*** sonali5 has joined #openstack-trove23:48
*** shayneburgess has joined #openstack-trove23:56
*** shayneburgess has quit IRC23:56

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