Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2013-12-12

| Channels | #fuel index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
23:00 e0ne joined #fuel
23:13 teran joined #fuel
23:15 mrasskazov joined #fuel
23:34 mrasskazov joined #fuel
23:42 mrasskazov joined #fuel
23:55 teran joined #fuel
01:22 Vidalinux joined #fuel
01:27 IlyaE joined #fuel
01:35 fandikurnia01 joined #fuel
01:36 dnovakovskiy joined #fuel
01:41 fandikurnia01 joined #fuel
01:48 wputra joined #fuel
01:52 rongze joined #fuel
01:59 wputra hi all
01:59 wputra i use fuel since 3.1, and now try 3.2.1 version
02:00 wputra it is possible to have multiple floating ip in environment that created by fuel 3.2.1?
02:01 wputra i already test ini fuel 3.2 before, i just add subnet ini predefined external network
02:02 wputra but now i want the new network that i allocate as floating IP separated by vlan
02:03 wputra yesterday i try trunks instead of tags in br-ex, so i can have several vlan in br-ex
02:04 wputra but it seems have negative effect to mirantis openvswitch cluster
02:05 wputra maybe any other approach to get the goal?
02:15 xarses joined #fuel
02:29 rongze joined #fuel
02:39 vkozhukalov joined #fuel
02:47 _ilbot joined #fuel
02:47 Topic for #fuel is now Fuel for Openstack: http://fuel.mirantis.com/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
02:48 teran joined #fuel
03:06 tonyha joined #fuel
03:07 rmoe joined #fuel
03:12 ykotko joined #fuel
03:12 alex_didenko joined #fuel
03:12 bogdando joined #fuel
03:27 ArminderS- joined #fuel
03:33 rongze_ joined #fuel
04:30 tonyha_ joined #fuel
04:33 dnovakovskiy joined #fuel
04:34 dnovakovskiy joined #fuel
05:12 angdraug joined #fuel
05:13 Shmeeny joined #fuel
05:17 SergeyLukjanov joined #fuel
05:20 rmoe joined #fuel
05:22 rongze joined #fuel
05:36 teran joined #fuel
05:41 AndreyGrebenniko joined #fuel
05:47 Fecn joined #fuel
05:51 SergeyLukjanov joined #fuel
06:01 teran joined #fuel
06:05 fandikurnia01 joined #fuel
06:20 rongze joined #fuel
06:36 rvyalov joined #fuel
07:09 e0ne joined #fuel
07:10 mihgen joined #fuel
07:26 dnovakovskiy hey guys
07:26 dnovakovskiy mirantis openstack 3.2.1. Is it mandatory for fuel master node to have access to internet?
07:26 dnovakovskiy i have a customer whose server box hangs after reboot upon completion of fuel master node installation - https://dl.dropboxusercontent.com/u/5844282/IMG_20131211_103111.jpg
07:26 dnovakovskiy nothing happens after rhsmcertd load
07:26 dnovakovskiy var/log/messages shows:
07:26 dnovakovskiy Dec 12 01:44:10 fuel kernel: sd 8:0:0:0: [sdb] Assuming drive cache: write through
07:26 dnovakovskiy Dec 12 01:44:10 fuel kernel: sdb: sdb4
07:26 dnovakovskiy Dec 12 01:44:10 fuel kernel: sd 8:0:0:0: [sdb] Assuming drive cache: write through
07:26 dnovakovskiy Dec 12 01:44:10 fuel kernel: sd 8:0:0:0: [sdb] Attached SCSI removable disk
07:26 dnovakovskiy Dec 12 01:44:31 fuel ntpd_initres[1689]: host name not found: 0.centos.pool.ntp.org
07:26 dnovakovskiy Dec 12 01:44:31 fuel ntpd_initres[1689]: host name not found: 1.centos.pool.ntp.org
07:27 dnovakovskiy Dec 12 01:44:31 fuel ntpd_initres[1689]: host name not found: 2.centos.pool.ntp.org
07:27 dnovakovskiy the box has no access to internet, that's why ntpd cant get to centos ntp servers
07:27 dnovakovskiy i'm thinking what to recommend - A) reconfigure centos to use local ntp OR B) tell customer to let fuel master node use internet (if something else but ntpd will require it later)
07:28 mrasskazov joined #fuel
07:44 nurla1 joined #fuel
07:45 mihgen dnovakovskiy: hi
07:45 mihgen it should work without access to Internet
07:46 dnovakovskiy thanks
07:46 SebastianKreutzb left #fuel
07:46 dnovakovskiy any other ideas why load may hang after rhsmcertd? current suggestion is that ntp does it
07:46 mihgen so the issue must be in somewhere else
07:47 mihgen do you have diagnostic snapshot from that installation?
07:47 dnovakovskiy at least ntd messages are the last ones in messages log...
07:47 dnovakovskiy nope. fuel master node doesn't boot after installation from iso
07:47 dnovakovskiy so no access to web ui
07:48 mihgen oh sorry I thought about other servers
07:48 mihgen it might be issue with disks
07:48 mihgen we really need all logs from the node, can you pack them from /var/log ?
07:48 mihgen and file ticket in jira
07:53 dnovakovskiy ok, will do
07:53 dnovakovskiy if ntp hook won't help
07:54 dnovakovskiy hmm
07:54 dnovakovskiy actually looks like i have them :)
07:54 dnovakovskiy https://dl.dropboxusercontent.com/u/5844282/fuellog.7z
07:57 SteAle joined #fuel
08:06 rongze joined #fuel
08:07 mihgen dnovakovskiy: I won't be able to look at this right away. Please file bug in Jira (we file customer's bugs there, if it's open project  / issue / doesn't contain sensitive info and customer's name - you can file to launchpad)
08:07 dnovakovskiy I prefer Jira. will do, thanks!
08:28 vkozhukalov joined #fuel
08:37 nurla1 joined #fuel
08:44 teran joined #fuel
08:45 anotchenko joined #fuel
08:46 teran_ joined #fuel
08:57 mihgen wow we've got a build for fuel library tests? https://fuel-jenkins.mirantis.com/job/fuellib_review_systest_centos/
09:01 mattymo syntax and full build? that's awesome
09:02 mattymo nurla, did you do this? ^^
09:02 nurla O:-)
09:34 ArminderS- joined #fuel
09:40 e0ne joined #fuel
09:51 anotchenko joined #fuel
09:51 miguitas joined #fuel
09:51 mrasskazov joined #fuel
10:17 SergeyLukjanov joined #fuel
10:23 anotchenko joined #fuel
10:32 e0ne_ joined #fuel
10:36 anotchenko joined #fuel
10:41 rongze joined #fuel
11:01 Fecn So.. I've been testing out the HA abilities of the controllers today.. (I have 3 of them)... I rebooted one of them to see what would happen, and not I can't get it to show back as 'online' in the output of crm_mon
11:01 Fecn Is there any special procedure needed to get it back after a reboot?
11:01 Fecn (I assumed that all the services would start up automatically)
11:26 anotchenko joined #fuel
11:40 teran joined #fuel
12:02 mihgen joined #fuel
12:06 anotchenko joined #fuel
12:09 xdeller joined #fuel
12:27 Fecn1 joined #fuel
12:28 anotchenko joined #fuel
12:34 teran_ joined #fuel
12:37 rongze joined #fuel
13:02 Fecn joined #fuel
13:11 rongze joined #fuel
13:20 anotchenko joined #fuel
13:23 SergeyLukjanov joined #fuel
13:25 MiroslavAnashkin joined #fuel
13:39 teran joined #fuel
13:41 anotchenko joined #fuel
13:48 MiroslavAnashkin joined #fuel
14:01 rongze_ joined #fuel
14:20 rongze joined #fuel
14:24 anotchenko joined #fuel
14:42 rongze_ joined #fuel
15:22 mattymo|wfh joined #fuel
15:23 Shmeeny joined #fuel
15:28 IlyaE joined #fuel
15:40 anotchenko joined #fuel
15:47 mihgen joined #fuel
16:02 jhurlbert joined #fuel
16:32 dnovakovskiy joined #fuel
16:35 rvyalov joined #fuel
16:41 ArminderS- joined #fuel
16:46 albionandrew joined #fuel
16:48 davideaster joined #fuel
17:04 angdraug joined #fuel
17:14 albionandrew joined #fuel
17:14 SergeyLukjanov joined #fuel
17:24 nurla1 joined #fuel
17:35 jhurlbert joined #fuel
17:39 rmoe joined #fuel
17:41 xarses joined #fuel
17:44 albionandrew Hi xarses - The disk issue I think is fixed. But we fail network verification now. http://picpaste.com/Screen_Shot_2013-12-12_at_10.41.55_AM-nMh9jz6q.png I have checked that the devices map to the private VLAN 101-130 on the flex 10 which they do... I don't think we have changed anything since our previous deployment. What next?
17:48 MiroslavAnashkin albionandrew: have you ran the failed network check against already deployed environment?
17:50 teran joined #fuel
17:50 albionandrew MiroslavAnashkin We had a working deployment, network configured etc and then moved to 3.2.1
17:56 rvyalov joined #fuel
17:56 MiroslavAnashkin albionandrew: Did you mean you upgraded just the master node, but preserved environment?
17:56 albionandrew No new Master node, new cluster. but left the settings on the hardware the same
18:01 MiroslavAnashkin albionandrew: I just wanted to say, network verification works correct only before the environment is deployed.
18:02 albionandrew I see.
18:02 aleksandr_null joined #fuel
18:02 albionandrew Not deployed new env yet so I guess it's failing somewhere?
18:03 albionandrew I rebooted and verification hasn't failed again yet. Green ants, how long should it take to complete?
18:04 albionandrew I can't remember does it say something like "success"?
18:06 MiroslavAnashkin albionandrew: yes, there is success message
18:07 albionandrew Thanks it hasn't failed again yet but I don't see a success  message either.
18:07 albionandrew Does it have a timeout?
18:08 MiroslavAnashkin albionandrew: Sometimes it take up to 15 minutes to rediscover the nodes. Or you may reboot all the nodes from restarted master node.
18:08 angdraug joined #fuel
18:10 e0ne joined #fuel
18:10 albionandrew http://pastebin.com/ke8wrh6t
18:12 vkozhukalov joined #fuel
18:14 MiroslavAnashkin Please run `mco ping` on master node and check if all the nodes available
18:14 MiroslavAnashkin `mso ping -v1`
18:15 MiroslavAnashkin Sorry, `mco ping -v`
18:18 xarses albionandrew: is this deployed yet?
18:18 albionandrew No
18:18 jhurlbert joined #fuel
18:19 xarses albionandrew: hmm, it should just be using native tags to check everything, I'd expect it to work
18:19 xarses albionandrew: also, you probably need to check the vlan splinters checkbox on the settings page, it won't help the pre-deploy network check though
18:21 albionandrew xarses was that what you did with dhblaz I remember something about that?
18:22 xarses albionandrew: correct
18:23 albionandrew So if I do that I should ignore the network check? Deploy then run health checks?
18:25 mihgen MiroslavAnashkin:  >> I just wanted to say, network verification works correct only before the environment is deployed. <- any bug on this? I never heard this
18:27 MiroslavAnashkin mihgen: Under Ubuntu.
18:28 mihgen what under ubuntu? bug please
18:29 MiroslavAnashkin mihgen: and with Neutron. There are blueprints, not bugs. It is missing functionality, at least Dima claims so.
18:30 mihgen ok. I know about Neutron. nova-network should work fine in both ubuntu & centos
18:31 MiroslavAnashkin mihgen: We've broke network check for Ubuntu recently. Expecting the fix tomorrow.
18:31 mihgen so can I know the bug # ?
18:32 mihgen and you mean we broke in master?
18:33 e0ne joined #fuel
18:35 dnovakovskiy joined #fuel
18:38 AndreyDanin mihgen, https://bugs.launchpad.net/fuel/+bug/1259935
18:40 AndreyDanin albionandrew, so what about "mco ping -v" from master node?
18:41 dnovakovskiy joined #fuel
18:42 jhurlbert joined #fuel
18:46 mihgen AndreyDanin: should not we keep it open unless package is pushed to mirror?
18:47 albionandrew xarses - Just talked to dhblaz. The failure is happening because untagged traffic where does clan splinters come into that?
18:48 AndreyDanin mihgen, I changed statis to 'in progress'
18:48 AndreyDanin mihgen, I changed status to 'in progress'
18:48 xarses vlan splinters only comes in to play during the puppet install because of Open Virtual Switch
18:52 Fecn1 joined #fuel
19:20 e0ne joined #fuel
19:20 mrasskazov joined #fuel
19:27 dan_a joined #fuel
19:47 mrasskazov joined #fuel
19:47 SergeyLukjanov_ joined #fuel
19:48 SergeyLukjanov joined #fuel
20:07 IlyaE joined #fuel
20:09 dnovakovskiy joined #fuel
20:23 rvyalov joined #fuel
20:46 albionandrew xarses the network verification is still running. Do I have to remove from postgres? If so is there a table I need to look in? ... sudo -u postgres psql then .... ?
20:47 xarses albionandrew: I'm not sure how to stop it. I think it spawns a process that runs in the background
20:47 xarses but im not sure
20:48 albionandrew A restart on the master? Or do you think I'm better off starting again?
21:08 dnovakovskiy joined #fuel
21:25 IlyaE joined #fuel
21:45 SergeyLukjanov joined #fuel
21:53 albionandrew xarses do we need to pass untagged packets for the check to work? dhblaz thinks the test fails because no default vlan set. Does that seem plausible ? Every node fails with an error like http://picpaste.com/Screen_Shot_2013-12-12_at_2.51.35_PM-dEuBOInz.png Did the test change? We have had this pass in the past? I have not changed the network config. The Vlan failing is the private vlans in our case 101-130.
21:58 Vidalinux joined #fuel
22:05 Shmeeny_ joined #fuel
22:21 xarses albionandrew: I'm not familiar with the network tests, other than they are frequently being updated to try to cover cases better. It's possible that they will pass if you specify a default vlan on the switch. If so please raise it as a bug, only interfaces with un-tagged network roles (admin network for example) should be running a test that requires that.
22:22 albionandrew xarses Thanks

| Channels | #fuel index | Today | | Search | Google Search | Plain-Text | summary