Thursday, 2014-01-09

*** nowen_phone (~AndChat72@99-174-92-191.lightspeed.tukrga.sbcglobal.net) has joined #wikid00:08
*** Troy__ has quit (Ping timeout: 272 seconds)00:25
*** nowen has quit (Quit: Leaving.)01:06
*** nowen_phone has quit (Quit: Bye)01:41
*** rfxn has quit (Ping timeout: 276 seconds)07:05
*** Teck7 (~teck7@MTRLPQ5401W-LP140-01-2925087393.dsl.bell.ca) has joined #wikid07:06
*** KORG|2 (~kvirc@crytek.dream.net.ua) has joined #wikid07:17
*** KORG has quit (Read error: Connection reset by peer)07:17
*** rudy6 (~rudy6@213.132.115.194) has joined #wikid12:47
*** nowen (~nowen@50-194-249-125-static.hfc.comcastbusiness.net) has joined #wikid13:33
*** rudy6 has quit (Ping timeout: 264 seconds)14:16
*** rudy6 (~rudy6@213.132.115.194) has joined #wikid14:17
*** Troy__ (329b98a8@gateway/web/freenode/ip.50.155.152.168) has joined #wikid15:02
nowenmorning Troy__15:02
Troy__morning Nick.. how did the basketball game go?15:03
nowengood, we won.  my son scored on a free throw too.15:04
nowenhe's 11. ;-)15:04
Troy__good deal.. my kids are a bit younger.. but it's coming soon15:04
nowenmixed bag in some ways.  if you're not careful activities will take over and actual family time can suffer15:05
Troy__Vince is wanting to move the db from the default /var/lib/pgsql/ to /data15:05
nowenhmm, why?  is it a special partition with extra space?15:05
Troy__is that an easy thing to do or is it more pain than it's worth?15:05
Troy__brb15:06
Troy__back.. yes.. i guess it's a special partition with extra space15:16
Troy__i think I convinced Vince to keep it on the default location for now and move it later(if possible)15:19
nowenIt looks like a simple change, but any rpm update would overwrite it15:22
Troy__ok.. so we'd have to modify the conf each time we update?15:26
nowenI think it is only in postgresql.conf, so only when that file changes15:26
nowenwhich isn't often15:26
nowenwhy does he want to change it?15:27
Troy__i think it's just how they have other dbs in the company setup..   separate partitions where it's easy to modify, backup, etc15:29
nowenwell, that may outweigh the risks.15:32
Troy__ok.. we'll see about switching in the lab once we have it running in the default location first15:35
nowenmakes sense15:35
Troy__so the SJ db is around 700MB right now15:35
nowenwhat's the status in the lab?15:35
Troy__my plan is to backup this DB, run a log archive, vacuum, then see if I can sync the db to the old primary again15:36
nowensounds good. should be a lot easier with a smaller db.15:37
Troy__the lab is still down.. I guess Vince is still working on getting the pgsql conf worked out15:37
nowenmaybe he should just start fresh.  seems like a worm hole15:38
Troy__yea. i agree15:38
Troy__another question I thought about for upgrading wikid15:39
nowenok15:39
Troy__when we get the lab up and running.. can I re-create all the certificates in the lab with the upgraded WiKID, then transfer them over to the production box (after upgrading) ?15:40
Troy__I hope that made sense15:40
Troy__i can't have the production down for more than a few minutes.. so I was thinking about creating them over on the lab side with the upgraded version15:41
nowenhmm, it might be better to just copy them from production to the lab.15:42
nowenoh wait15:42
nowenI'm with you now.15:42
Troy__i'm just talking about upgraded WiKID15:42
nowenlet me double check15:42
*** rudy6 has quit (Quit: Nettalk6 - www.ntalk.de)15:42
Troy__for each network client ( and I have 7 or 8), I need to create a cert and copy over to each of my SAML boxes15:43
Troy__now that I think about it.. I don't think that would save me much time.. because I would still have to delete and re-add the network certificates one at a time in production15:46
nowenit should work - create the network clients in the lab15:47
nowenyou think if you get the lab setup, you'll be able to do upgrades more frequently?15:48
Troy__yes. i sure hope so15:52
nowenlike to update postgresql15:52
Troy__the main thing that's holding me back from upgrading to the latest version is the new inter certification and replication15:53
Troy__upgrading in replication mode is more complicated15:53
Troy__because you have to upgrade both servers before you can resume replication15:54
nowenyeah15:55
nowendoes Vince have a preferred version of pgsql?  we can test it15:56
Troy__I'll ask him15:58
Troy__one feature I think would be nice to add in a future revision is the ability to clean up devices from the WiKID admin panel15:59
nowenwe've got lab capability to test everything you're doing too, so we can guide you to minimize downtime15:59
Troy__such as remove devices with no activity greater than 1 year16:00
Troy__ok16:00
nowengood one16:00
Troy__maybe a bit dangerous16:00
nowencould be, but also could add security16:01
nowenremove from db or just display?16:01
Troy__Vince says they are using 9.3 .. but he said the latest version of 8 would work but doesn't know if it's supported much anymore16:03
nowenok16:03
Troy__i think having a button that the admin could choose either remove from db or disable (and hide from view)16:04
Troy__we have almost 16,000 devices right now and it's getting hard to manage16:05
Troy__we don't have any easy way to sort or clean this up easily16:05
nowenyeah16:05
nowenI think we could do a filtered page more like the logs.16:05
Troy__we need to upgrade before the sorting works :(16:05
Troy__yea.. that would be awesome16:05
nowenoh - you're pre that fix?16:06
Troy__yea.. i brought that to you about a year back i think16:06
Troy__we haven't upgraded since16:06
nowenyeah - looking at the change log16:07
Troy__i remember you fixed in the next version16:07
Troy__Vince is saying we should just go to the latest PGSQL 9 if we're going through the trouble of upgrading anyway16:08
nowenok - going to test it16:08
Troy__You probably just saw the latest dbsize email17:49
Troy__it was after we archived and ran a vacuum17:50
nowenmuch better17:50
Troy__i may try and sync the db back over to the old primary later this afternoon17:52
nowenok17:52
*** Troy__ has quit (Ping timeout: 272 seconds)19:44
*** Troy__ (329b98a8@gateway/web/freenode/ip.50.155.152.168) has joined #wikid20:05
Troy__finally got our lab back online again20:07
Troy__we are actually running a newer build in the lab than what we are running in production20:08
Troy__not sure how that happened20:08
Troy__lab = 3.5.0-b1411 / production = 3.4.87-b121620:10
Troy__any difference in db schema with those 2 revisions?20:10
nowenlet me see20:16
nowenyes20:29
nowenyou can see yourself by running " more /opt/WiKID/conf/database/migrations/cumulative.sql" on both machines20:29
nowen18 is the latest rev20:30
nowenTroy__: I also have it working on postgresql 9.321:22
Troy__good deal21:23
nowenyou have to do a dump and restore21:23
nowenI installed it beside 8.1 and then upgraded21:23
nowenI can document how I did it, but Vince might have his own preferences or corp standards21:30
Troy__ok. that sounds good.. we plan to fail back over to our old primary server tomorrow afternoon21:45
Troy__we can test the postgresql upgrade in the lab now that it's up and running21:45
*** nowen has quit (Quit: Leaving.)23:17
*** Troy__ has quit (Ping timeout: 272 seconds)23:54

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!