Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel-dev, 2014-05-27

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

All times shown according to UTC.

Time Nick Message
01:55 teran joined #fuel-dev
04:28 IlyaE joined #fuel-dev
04:48 e0ne joined #fuel-dev
04:49 e0ne joined #fuel-dev
05:17 IlyaE joined #fuel-dev
06:04 IlyaE joined #fuel-dev
06:25 sambork joined #fuel-dev
06:53 artem_panchenko joined #fuel-dev
07:06 salmon_ joined #fuel-dev
07:35 IlyaE joined #fuel-dev
07:57 teran joined #fuel-dev
08:08 ios77 joined #fuel-dev
08:10 ios77 joined #fuel-dev
08:17 brain461 joined #fuel-dev
08:21 EeeBotu New bug #1323548 in fuel: "HA: 3th controller went to 'ERROR' state during the deployment" [High, Confirmed] https://bugs.launchpad.net/bugs/1323548
08:49 EeeBotu New bug #1323555 in fuel: "Storage settings in Fuel UI" [Undecided, New] https://bugs.launchpad.net/bugs/1323555
09:07 teran_ joined #fuel-dev
09:35 tatyana joined #fuel-dev
09:54 zynzel guys, any plan to update corosync to ver 2.x in fuel?
09:54 teran joined #fuel-dev
09:54 holser zynzel, we have plans
09:55 holser and refactor manifest to use pcs instead of crm
09:55 zynzel holser: any 5.0? 5.1?
09:55 zynzel not decided?
09:55 holser 5.1 - 6.0
09:55 holser most likely 5.1
09:57 zynzel k, thx
09:57 holser zynzel, I alose saw your questions in #linux-ha, feel free to ask me, bogdando, xenolog
09:57 zynzel holser: im writing new RA for mysql
09:57 holser zyntel, it’s in plan for 5.1
09:58 zynzel but i use crm_attribute to store gtid of each galera cluster, which is updated in mysql_monitor()
09:58 holser so we may join forces to speed up it
09:58 zynzel but when fuel start puppet and try to configure resource i have database diff error, because my RA update pacemaker configuration
09:58 zynzel "outside" puppet...
09:59 holser I am not sure if your approach is correct, it would be better to ask xenolog as he is designing how it should behave
10:00 zynzel holser: well, i have everything ready - i only need to store GTID in some way, which will not destroy fuel puppet manifests
10:00 zynzel support restarting whole cluster, single node, boostrap etc.
10:00 zynzel xenolog: ayt?
10:02 zynzel holser: do you know if we have any blueprint for galera RA?
10:03 holser We’ll have a meeting tomorrow where we plan features for next release
10:04 holser also we have document where we ask what we would like form next release, what service teams need
10:04 holser once we plan, we’ll start working on BPs
10:15 mattymo we need to switch to xtrabackup instead of rsync for galera
10:15 mattymo it will improve performance a lot
10:15 mattymo also in the newest galera from percona they support a special preferred donor mode that will also improve recovery time
10:15 holser mattymo, we need to switch from home made galera to percona
10:15 mattymo I think in the vanilla galera too
10:16 mattymo you can do node-1 is master, node-2 and node-3 are preferred donors. if 1 dies, 2 can become master (and drop donor role), then 3 can be a donor to 1 when it recovers
10:16 mattymo that way you can still write to the DB and recover a failed node
10:16 zynzel mattymo: RA is independent from galera below :)
10:16 mattymo RA?
10:16 zynzel we can use percona, galera, or mariadb.
10:17 akasatkin mihgen: was my path reverted in 5.0-26 ? https://bugs.launchpad.net/fuel/+bug/1323433  seems to be reproducible on it.
10:17 zynzel resource agent for pacemaker.
10:17 holser it would be interesting to compare percona with maria
10:17 mattymo you can run percona on top of maria as well
10:17 zynzel mattymo: but what happend if node 3 have newest version of DB and not node-1 or node-2?
10:17 zynzel RA should solve this.
10:17 mattymo that's only important in the case where all 3 nodes get shut down
10:18 tnurlygayanov Hi team
10:18 mattymo galera can resolve itself if 3 is newer than 1 and 2, but not when it's unsure who is the master
10:18 mattymo good day tnurlygayanov
10:18 tnurlygayanov I have some strange bug with my Fuel environment
10:19 tnurlygayanov can anybody help to find the root of problem?
10:19 tnurlygayanov Deployment has failed. Error occurred while running method 'deploy'. Inspect Orchestrator logs for the details.
10:19 tnurlygayanov I can provide the access to my Fuel environment
10:20 mattymo tnurlygayanov, look at orchestrator logs in fuel ui :)
10:21 tzn joined #fuel-dev
10:27 EeeBotu New bug #1323581 in fuel: "Huge environment tests missed mongo role with deployed ceilometer" [Medium, In Progress] https://bugs.launchpad.net/bugs/1323581
10:38 teran joined #fuel-dev
10:45 themistymay joined #fuel-dev
10:47 sambork joined #fuel-dev
10:59 e0ne joined #fuel-dev
11:01 tnurlygayanov hm, how I can get the list of IP addresses of all slave nodes?
11:01 tnurlygayanov I lost 5 nodes from my 50-nodes cluster and I want to detect the IP addresses of these nodes
11:07 themistymay joined #fuel-dev
11:07 EeeBotu New bug #1323597 in fuel: "Remove merge_array function" [Medium, In Progress] https://bugs.launchpad.net/bugs/1323597
11:09 zynzel holser: xenolog https://lists.launchpad.net/fuel-dev/msg01100.html
11:10 zynzel mihgen: https://lists.launchpad.net/fuel-dev/msg01100.html
11:10 zynzel :)
11:34 EeeBotu New bug #1323608 in fuel: "External gateway seems down in horizon, but works" [Undecided, New] https://bugs.launchpad.net/bugs/1323608
11:34 EeeBotu New bug #1323607 in fuel: "vbox_scripts for fuel 4.1 don't create VDI image for fuel master under cygwin" [Undecided, New] https://bugs.launchpad.net/bugs/1323607
11:34 EeeBotu New bug #1323600 in fuel: "Measure system test thread loads for every microcloud type" [High, Confirmed] https://bugs.launchpad.net/bugs/1323600
11:38 al_ex joined #fuel-dev
11:42 death_ joined #fuel-dev
11:49 e0ne joined #fuel-dev
12:09 IvanBerezovskiy joined #fuel-dev
12:13 Loic__ joined #fuel-dev
12:16 Loic__ hello guys, I'm working on fuel customize for Orange in France and I encounter a problem, maybe it's a knew one and you can help me . From yesterday all my deployments failed after the installation of the ubuntu OS.
12:16 e0ne joined #fuel-dev
12:17 Loic__ I find in mcollective log some warning 11:59:36.604840 #1040]  WARN -- : netio.rb:391:in `_init_line_read' PLMC7: Exiting after signal: SignalException: SIGTERM
12:17 Loic__ did you know if it's possible I have some corrupted/inconsistency data somewhere in fuel ?
12:25 mattymo Loic__, did you make any changes to the original Fuel codebase?
12:25 Loic__ I make change on puppet script
12:26 mattymo Loic__, this warning is not a problem. It would be better if you could get us a little closer to what is failing. Do nodes start running puppet? Are there puppet-apply.log files in your /var/log/remote directory?
12:26 mattymo Loic__, there may be a syntax or conflict in your Puppet changes that broke deployment.
12:27 mattymo syntax error*
12:27 mattymo usually this finds errors in Puppet: grep -R err: /var/log/remote
12:28 Loic__ On my controller node puppet didn't start, I have no log, and on my compute node I have puppet log with no special error
12:33 Loic__ with grep -R err: /var/log/remote I find some err like this :
12:34 Loic__ "/var/log/remote/10.20.0.6/ovs-vsctl.log:2014-05-27T12:20:40.332021+00:00 err:  00001|vsctl|ERR|no port named eth0"
12:35 Loic__ and other I'll try to find why :)
12:36 mattymo aglarendil, dilyin_ https://review.openstack.org/#/c/91930/ is ready to review and merge
12:36 mattymo Loic__, no port named eth0 sounds terrible
12:36 mattymo are you using strange hardware?
12:36 Loic__ no it's a vmware vm
12:36 mattymo we have issues with a few virtualized platforms and interface types with openvswitch
12:37 mattymo MiroslavAnashkin, is not at his desk at the moment, but I remember him saying one particular virtualized card is better for running ontop of VMWare
12:37 mattymo on top*
12:37 dilyin_ [actually modern linuces don't have eth0 at all]
12:37 mattymo he should be back in a moment
12:37 mattymo dilyin_, we're not modern :)
12:38 Loic__ ok
12:38 Loic__ did you think I can launch the puppet manifest manually on node for checking error ?
12:39 mattymo Loic__, you have networking issues that probably came during the first steps of provisioning. You should reset your environment and try again. But you should power down the VMs and make sure all NICs are using this hardware:
12:39 mattymo Vlance: This is an emulated version of the AMD 79C970 PCnet32- LANCE NIC
12:39 mattymo this one works best with openvswitch
12:39 Loic__ ok thansk I'll try that so
12:40 mattymo you can try to run puppet manually, but the Fuel Orchestrator can't talk to it, so it will not report as a successful deployment
12:41 Loic__ the strange point is that deployment works last week ...that's why i first think to corrupted data because I reset envinoment many time for test
12:41 MiroslavAnashkin Loic__: Please check virtual machine settings file, (.vmx), the following settings:
12:42 MiroslavAnashkin 1. Use 'vlance' network card if possible. e1000 works in selected VMware versions (mostly enterprise ones) and may not work with others.
12:42 MiroslavAnashkin You may open .VMX file of your machine (OK, all machines) and check the following string for each network adapter:
12:42 MiroslavAnashkin `ethernetXXX.virtualDev = "vlance"`.
12:42 MiroslavAnashkin If there is no vlance (old AMD) support in your distribution - use e1000.
12:42 MiroslavAnashkin 2. Allow promiscuous mode for each network card. Again, check VMX file of your VM, the following string:
12:42 MiroslavAnashkin `ethernetXXX.noPromisc = "false"`
12:42 MiroslavAnashkin 3. Check you network card PXE boot is not turned off  (affects all cards that use selected driver and should be on by default).
12:42 MiroslavAnashkin `vlance.noOprom = "false"`
12:42 MiroslavAnashkin 4. In case you use ESXi/vSphere with vSwitch as networking solution:
12:42 MiroslavAnashkin Create separate port group or even separate vSwitch instance with port group for each virtual network.
12:42 MiroslavAnashkin Allow promiscuous mode for each port group.
12:42 MiroslavAnashkin Please note: some kinds of multicast/broadcast traffic does not go outside the port group, so always use port groups for Neutron.
12:42 MiroslavAnashkin vSwitch does not support Native VLAN for Trunk ports and removes all VLAN tags from traffic, going through Access ports. So you cannot mix VLAN tagged and untagged traffic and have to create different port groups for Trunk and Access ports.
12:43 * holser hides from MiroslavAnashkin
12:44 Loic__ ok , thanks for all informations ! :)
12:48 rmoe joined #fuel-dev
12:55 mattymo rmoe, you are here early, sir
13:22 jaypipes joined #fuel-dev
13:23 salmon_ hmm, when should we send mail to fuel-dev and when to openstack-dev?
13:29 mattymo when the mood is right
13:56 IlyaE joined #fuel-dev
14:02 tnurlygayanov hm... ISO with 4.1.1 shown incorrect version of OpenStack
14:03 tnurlygayanov now it is OpenStack Havanna 2013.2.2, but looks like we already has 2013.2.3
14:03 tnurlygayanov probably this parameter just hardcoded in MOS UI?
14:03 tnurlygayanov or it means that we have incorrect ISO with 4.1.1?
14:05 MiroslavAnashkin What ISO 4.1.1 version? We have 4.1A, 4.1B, several versions of each of these tqo.
14:05 MiroslavAnashkin s/tqo/two/
14:05 tnurlygayanov 327 ISO
14:06 tnurlygayanov the latest
14:08 b-zone joined #fuel-dev
14:13 tnurlygayanov ok, I will describe this issue and if it is incorrect issue, we can mark it as 'Invalid'.
14:13 tnurlygayanov but looks like it is small bug in UI or big problem with ISO building proccess.
14:13 tnurlygayanov because I want to verify the exact version of OpenStack
14:19 teran joined #fuel-dev
14:37 EeeBotu New bug #1323676 in fuel: "Update puppetlabs-stdlib to 4.x.x" [Low, New] https://bugs.launchpad.net/bugs/1323676
14:41 mattymo holser, fail ^
14:42 holser mattymo, why?
14:42 mattymo https://bugs.launchpad.net/fuel/+bug/1253645 is from november)
14:42 mattymo I marked your new bug as a duplicate
14:42 designated is there a specific time 5.0 will be released today or is there a delay?
14:42 holser okey, but 4.x fully supports puppet 2.7
14:42 mattymo it's ok just it's a duplicate bug
14:43 holser here is a compatibility matrix https://github.com/puppetlabs/puppetlabs-stdlib
14:43 holser okey, thank you matt
14:44 holser I spent a hour testing it, it works fine with our current puppet 3.4.x
14:45 mattymo it's more a matter of making sure we aren't using any deprecated syntax
14:46 mattymo make sure deployment works fine in HA and non-HA
14:46 mattymo I doubt it changed much, but it's always a concern
14:46 holser yeah, your are right
14:46 holser we’ll test it thoroughly. I am sure
14:47 mattymo just put up an iso on jenkins and I can do a test too
14:47 mattymo split up ubuntu/centos and get it covered effectively
14:47 holser okey… thanks a lot
14:50 MiroslavAnashkin designated: There is no specific time.
14:50 designated MiroslavAnashkin, right on
15:06 designated joined #fuel-dev
15:09 IvanBerezovskiy left #fuel-dev
15:32 tnurlygayanov bug with OpenStack version for 4.1.x: https://bugs.launchpad.net/fuel/+bug/1323723
15:37 dshulyak_ joined #fuel-dev
15:41 EeeBotu New bug #1323723 in fuel: "Incorrect version of OpenStack Havanna release in MOS 4.1.1" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1323723
15:41 EeeBotu New bug #1323707 in fuel: "Old lvm metadata on partitons breaks deployment of ceph-osd" [Medium, In Progress] https://bugs.launchpad.net/bugs/1323707
15:41 EeeBotu New bug #1323705 in fuel: "Instance console can't connect on the first attempt" [Undecided, New] https://bugs.launchpad.net/bugs/1323705
15:44 tnurlygayanov MiroslavAnashkin, {"build_id": "2014-05-27_04-18-31", "mirantis": "yes", "build_number": "327", "nailgun_sha": "e5e2894703bd5e5218ad96d6884bd39dc8e4e921", "ostf_sha": "3690709d95c1f42506b6d955b44221910cec8527", "fuelmain_sha": "2ff157e46f5ee5920280e60b8d4ff6140d46183c", "astute_sha": "55df06b2e84fa5d71a1cc0e78dbccab5db29d968", "release": "4.1B", "fuellib_sha": "c64ffdc5e7bfe7fa8eb971be1fea2260ee849ab5"}
15:57 rmoe joined #fuel-dev
15:59 MiroslavAnashkin tnurlygayanov: Nova package  version from local repo on latest 4.1B ISO: openstack-nova-2013.2.3.fuel4.1.1-mira0.noarch.rpm
16:00 tnurlygayanov ok, cool. so, it is small bug in UI.
16:01 MiroslavAnashkin So, yes, it has updated OpenStack distro.
16:02 teran joined #fuel-dev
16:23 e0ne joined #fuel-dev
16:37 jobewan joined #fuel-dev
16:43 EeeBotu New bug #1323749 in fuel: "Arbitrary conflicting cluster settings can be set via .set() method of Settings model" [Low, Confirmed] https://bugs.launchpad.net/bugs/1323749
17:04 xwizard joined #fuel-dev
17:09 designated Trying to build ISO on ubuntu 12.04, followed the instructions here: http://docs.mirantis.com/fuel-dev/develop/env.html  but getting "make: *** [/build/docker/cobbler.done] Error 2".  Any ideas?
17:27 teran joined #fuel-dev
17:31 xarses joined #fuel-dev
17:34 holser designated, it would be nice to see your logs
17:35 holser just http://pastebin.com/ them so we’ll be able to investigate
17:37 designated holser, i did a make deep_clean now trying to build again.
17:44 teran joined #fuel-dev
17:55 e0ne joined #fuel-dev
18:49 IlyaE joined #fuel-dev
19:08 IlyaE joined #fuel-dev
19:24 angdraug joined #fuel-dev
19:58 BillTheKat joined #fuel-dev
20:14 angdraug joined #fuel-dev
20:34 IlyaE joined #fuel-dev
20:44 angdraug joined #fuel-dev
20:48 tzn joined #fuel-dev
21:11 e0ne joined #fuel-dev
21:21 e0ne joined #fuel-dev
21:31 openstackstatus NOTICE: Zuul is offline due to an operational issue; ETA 2200 UTC.
21:31 Topic for #fuel-dev is now Zuul is offline due to an operational issue; ETA 2200 UTC.
22:01 EeeBotu New bug #1323845 in fuel: "Fuel CI for Fuel Docs should include link to build results in gerrit comment" [Wishlist, New] https://bugs.launchpad.net/bugs/1323845
22:07 Topic for #fuel-dev is now #fuel-dev Fuel dev docs http://docs.mirantis.com/fuel-dev/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel-dev/ | gerrit traffic @ #fuel-tracker
22:08 openstackstatus NOTICE: Zuul is started and processing changes that were in the queue when it was stopped. Changes uploaded or approved since then will need to be re-approved or rechecked.
22:15 tzn joined #fuel-dev
23:01 xarses joined #fuel-dev
23:19 tzn joined #fuel-dev
23:58 teran joined #fuel-dev

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