Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2015-10-12

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

All times shown according to UTC.

Time Nick Message
00:27 sbog joined #fuel
00:28 vsedelnik joined #fuel
00:50 sbog joined #fuel
01:03 EinstCrazy joined #fuel
01:05 zhangjn joined #fuel
01:25 zhangjn_ joined #fuel
01:29 vsedelnik joined #fuel
01:47 ilbot3 joined #fuel
01:47 Topic for #fuel is now Fuel 7.0 (Kilo) https://software.mirantis.com | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
01:59 zhangjn joined #fuel
02:13 zhangjn joined #fuel
02:20 zhangjn joined #fuel
02:25 Sesso_ THey should be at the login until you deploy them
02:30 vsedelnik joined #fuel
02:33 lawrancejing joined #fuel
02:36 gongysh joined #fuel
02:45 lawrancejing joined #fuel
02:46 zhangjn joined #fuel
03:00 zhangjn joined #fuel
03:23 zhangjn joined #fuel
03:30 zhangjn joined #fuel
03:30 vsedelnik joined #fuel
04:08 zhangjn joined #fuel
04:09 zhangjn joined #fuel
04:43 subscope joined #fuel
04:50 vsedelnik joined #fuel
05:18 sbog joined #fuel
05:39 sbog joined #fuel
05:46 cartik joined #fuel
06:13 Sesso_ How do I change the SSL cert on my horizon dashboard if I dont have a openstack-dashboard.conf file. Is it pulling the current self signed cert from somewhere else? If so, can I create a openstack-dashboard.conf  so I can use my signed cert?
06:13 pradiprwt joined #fuel
06:29 cartik hi, can someone help me in pointingout the difference between the plugins 6.0 vs 7.0
06:29 cartik sorry 6.1
06:29 subscope joined #fuel
06:54 mkwiek joined #fuel
06:56 sbfox joined #fuel
07:12 pradiprw_ joined #fuel
07:15 vsedelnik joined #fuel
07:18 [HeOS] joined #fuel
07:29 neilus joined #fuel
07:34 aliemieshko_ joined #fuel
07:45 sergmelikyan joined #fuel
07:52 hyperbaba joined #fuel
07:53 subscope joined #fuel
07:55 subscope joined #fuel
08:05 meow-nofer joined #fuel
08:06 bdudko joined #fuel
08:06 nurla joined #fuel
08:06 mrasskazov joined #fuel
08:07 izinovik joined #fuel
08:07 mihgen joined #fuel
08:08 aglarendil_ joined #fuel
08:08 evg joined #fuel
08:09 xenolog joined #fuel
08:09 MorAle joined #fuel
08:09 alrick joined #fuel
08:10 vkramskikh joined #fuel
08:11 rvyalov joined #fuel
08:12 pradiprwt joined #fuel
08:12 apalkina joined #fuel
08:13 sbog joined #fuel
08:14 MiroslavAnashkin joined #fuel
08:15 akislitsky joined #fuel
08:20 JoeStack joined #fuel
08:21 subscope joined #fuel
08:24 e0ne joined #fuel
08:32 dklepikov joined #fuel
08:46 omolchanov joined #fuel
08:52 zhangjn joined #fuel
08:52 jaypipes joined #fuel
08:54 subscope joined #fuel
09:06 subscope joined #fuel
09:08 pradiprwt joined #fuel
09:25 aliemieshko_ joined #fuel
09:27 adanin joined #fuel
09:41 zhangjn joined #fuel
09:56 sergmelikyan joined #fuel
10:06 e0ne_ joined #fuel
10:11 Tatyanka_Leontov joined #fuel
10:14 aarefiev joined #fuel
10:16 aliemieshko joined #fuel
10:18 alrick joined #fuel
10:19 e0ne joined #fuel
10:21 cartik joined #fuel
10:36 bpiotrowski cartik: you mean what changes are needed, or…?
10:37 zerda joined #fuel
10:39 pradiprwt hi bpiotrowski, for me the changes i did in only metadata.yaml file and it's working fine ...Is there any other changes needed for fuel 7.0.?
10:39 bpiotrowski no, I don't think so, that's why I ask cartik for expanding his question
10:42 piontec joined #fuel
10:46 vsedelnik joined #fuel
10:51 pradiprwt bpiotrowski : Is there document is available for package version 3.0 ..?
10:52 pradiprwt i found there is some extra file are creating in fpb 3.0
10:52 az__ joined #fuel
10:56 sergmelikyan joined #fuel
10:57 zhangjn joined #fuel
10:58 EinstCrazy joined #fuel
10:59 bpiotrowski not sure about documentation, but example is rather fine: https://github.com/stackforge/fuel-plugins/tree/master/fuel_plugin_example_v3
11:00 EinstCrazy joined #fuel
11:00 bpiotrowski it's easier to write ha-enabled plugins now, as you can tell fpb about primary roles, and/or assign VIPs
11:01 bpiotrowski I guess that's the main change
11:03 vsedelnik joined #fuel
11:10 pradiprwt ok
11:12 alrick joined #fuel
11:12 sergmelikyan joined #fuel
11:12 alrick joined #fuel
11:16 pradiprwt bpiotrowski : while env deployment in fuel 7.0 , is necessary to add storage node..?
11:18 pradiprwt here my deployment got failed normally after adding storage node it's working fine ..
11:18 cartik joined #fuel
11:26 cartik sorry, was away for a while
11:26 bpiotrowski pradiprwt: no, it's not
11:26 cartik bpiotrowski, the question is to usderstand the change required to use he plugin developed in 6.1 to 7
11:27 bpiotrowski pradiprwt: please report a bug if it does, but storage node is not a requirement
11:31 bpiotrowski cartik: no idea, let me find someone who does…
11:31 bpiotrowski but I have a feeling that updating metadata should do
11:31 cartik that would be of much help, thanks
11:32 cartik will give that a try by updating the metadata
11:32 pradiprwt bpiotrowski : ok I will test it again if error come , will report the bug
11:40 subscope joined #fuel
11:40 sergmelikyan joined #fuel
11:41 bpiotrowski cartik: yeah, updating the metadata should do, there have been just new features added: https://wiki.openstack.org/wiki/Fuel/Plugins#7.0_features
11:41 bpiotrowski or so the friendly manual says
11:42 cartik bpiotrowski, Thanks, will give it a try by updating the metadata and update
11:44 vsedelnik joined #fuel
11:46 sergmelikyan joined #fuel
11:46 cartik bpiotrowski, Thanks it worked for me by just updating the metadata
11:48 bpiotrowski great
11:51 cartik joined #fuel
12:06 sergmelikyan joined #fuel
12:07 sergmelikyan joined #fuel
12:09 subscope joined #fuel
12:12 subscope joined #fuel
12:17 vsedelnik joined #fuel
12:19 e0ne joined #fuel
12:25 cartik joined #fuel
12:26 pradiprw_ joined #fuel
12:42 tkhno joined #fuel
12:42 sergmelikyan joined #fuel
12:45 DevStok joined #fuel
12:54 sergmelikyan joined #fuel
12:58 DevStok hi
12:58 tkhno joined #fuel
12:58 DevStok how many vlan are needed for private network?
13:04 DevStok can I set only two vlan?
13:10 sergmelikyan joined #fuel
13:14 pradiprwt joined #fuel
13:31 lawrancejing joined #fuel
13:32 DevStok annyone
13:37 mwhahaha Not sure i understand your question
13:37 mwhahaha what are you trying to do?
13:38 DevStok hi
13:38 DevStok setting kilo networks
13:38 DevStok private netwrok default values are 1000-1030
13:38 DevStok neutron l2 configuration
13:40 tatyana joined #fuel
13:40 mwhahaha ok?
13:41 DevStok nope
13:42 DevStok cause I need to understand which is the minimun number of vlan
13:42 DevStok that kilo needs
13:42 DevStok for private network
13:42 mwhahaha what network configuration did you setup? gre/vxlan/vlan?
13:43 DevStok vlan segmentation sorry
13:43 mwhahaha 3 i think
13:43 mwhahaha maybe 2 if you get creative
13:44 DevStok mmm
13:44 DevStok so i need 2 vlan for classic storage and mgmt networks
13:45 DevStok and other 3 vlan for private network
13:45 DevStok what kind of limitations can i'l face in the future using only 3 vlan ?
13:45 mwhahaha oh i thought you were asking how many vlans you need for the whole thing
13:46 mwhahaha you'd be restricting the number of tenant networks i beleive
13:46 mwhahaha it would be better to use gre/vxlan if you can
13:48 DevStok awh right
13:48 DevStok another question
13:48 DevStok i got 3 controller
13:48 DevStok mongo db has to be installed in all controllers right?
13:49 mwhahaha for ceilometer?
13:50 omolchanov joined #fuel
13:51 mwhahaha i don't think you need to install it on the controllers
13:51 mwhahaha i think you can dedicate specific nodes for it
13:52 DevStok on the compute neither?
13:52 mwhahaha no you'd probably want specific instances for it
13:52 mwhahaha for scaling/performance reasons
13:53 mwhahaha of course it depends on the size of your cloud
13:56 DevStok thanks again
13:56 DevStok and cinder storage is better to put on the compute?
13:57 ppradhan joined #fuel
13:58 mwhahaha probably not because if you lost a compute you might impact other vms on other computes
13:58 DevStok so is better to put on the controllers?
13:59 mwhahaha it really depends on the quality of your systems.  I guess it's a risk you have to figure out if you're comfortable with
14:00 mwhahaha you'll have competing IO on controllers
14:00 DevStok in previous env I put cinder on the computes
14:08 sergmelikyan joined #fuel
14:14 DevStok deployed
14:14 DevStok let's see
14:14 DevStok thansk again
14:36 sbog_ joined #fuel
14:40 sbog_ joined #fuel
14:46 jobewan joined #fuel
14:46 claflico joined #fuel
14:47 sergmelikyan joined #fuel
15:08 sergmelikyan joined #fuel
15:34 blahRus joined #fuel
15:42 goukos joined #fuel
15:45 goukos Hey guys, I'm quite new to OpenStack. I have deployed OpenStack on Fuel 6.1. How can I setup active/passive configuration for instances except for building a HA environment?
15:47 dklepikov what do you mean?
15:55 goukos dklepikov: I want to setup an active/passive configuration, by duplicating the instance data on multiple hosts. So, in case one of them fail, I can still have the instance duplicated on another host and I will not lose any data of the instance
15:57 sbog__ joined #fuel
15:58 dklepikov goukos: do you want to use LBaaS?
15:58 dklepikov goukos:  Am I right?
15:59 goukos dklepikov: not really, LBaaS is load balancing the requests.
16:00 dklepikov goukos: Please describe what result do you want to reach? "active/passive configuration for instances" what do you mean?
16:00 cartik joined #fuel
16:01 goukos dklepikov: let me rephrase. When you create an instance, let's say, is hosted on compute host A. I want to make a duplicate of that instance on compute host B and if the host A fails, the user can request his instance not on host A as he did, but on host B.
16:02 alrick joined #fuel
16:07 goukos dklepikov: except for the fail thing, the same concept applies if I want to update or put offline the host A, so I turn the instance on host B "active" and the user has access to his instance running on host B, not host A.
16:08 Zer0Byt__ joined #fuel
16:14 dklepikov goukos: do you want to use tis case for any instance (do you want to have a backup copy of any instance?)
16:17 goukos not exactly a backup copy, if you mean that this copy is offline. I want to boot the same image on two hosts at the same time, share the filesystem, share whatever the user does on the instance.
16:18 goukos and if I burn one host, the user has access to his data with no exception on the 2nd host
16:19 kozhukalov_ joined #fuel
16:38 dklepikov goukos: https://wiki.openstack.org/wiki/Cinder/blueprints/multi-attach-volume
16:40 dklepikov goukos: Looks like this functionality partly available started from Kilo release of the OpenStack
16:40 goukos dklepikov: yes but even if this was implemented it's only about the volume, not about the services running on the instance
16:41 goukos I want a copy of the whole state of the instance
16:41 serg_melikyan joined #fuel
16:41 Zer0Byt__ joined #fuel
16:43 dklepikov goukos: whole state of the instance with memory and all running services?
16:43 goukos yes
16:44 dklepikov goukos: If I remember correct, libvirt does not support it
16:45 sbog__ As I know, there is no way to do it by qemu itself. You can try to share filesystem on those VM's and setup some HA stuff like corosync and pacemaker on them manually
16:46 goukos I thought it would be kinda easy.. In my head sounds like an "endless" instance migration
16:53 goukos So, when I setup a HA environment on Fuel, the high availability is on what?
16:53 goukos I mean, if a compute host fails, the instances running on that host are off. That's not that HA :P
16:55 dklepikov goukos: you can use nova evacuate for instances
16:56 goukos as I read here https://ask.openstack.org/en/question/59964/openstack-instance-high-availabilityhow-to-make-pets-vm-highly-available/ the nova evacuate is not an automatic way of instance fail-over
16:56 dklepikov goukos: and use nova ephemeral volume for nova
17:23 scolby joined #fuel
17:29 scolby Hi all - I've deployed Fuel 7 on VirtualBox (Ubuntu 14.04 LTS) using the setup scripts.  Everything LOOKS ok, but I can't deploy even a simple environment because something (master node?) cannot talk to any repo - e.g. archive.ubuntu.com
17:31 scolby Any suggestions on what I might look for?  Thanks!
17:35 Zer0BYte__ what erro u get?
17:36 Zer0BYte__ scolby
17:52 xarses joined #fuel
17:56 Zer0Byt__ joined #fuel
18:24 sergmelikyan joined #fuel
18:31 sbog_ joined #fuel
18:42 Zer0Byte__ joined #fuel
18:50 scolby Zer0BYte:  Error  HTTPConnectionPool(host='archive.ubuntu.com', port=80): Max retries exceeded with url: /ubuntu/ (Caused by <class 'socket.gaierror'>: [Errno -3] Temporary failure in name resolution)
18:50 sbog_ Seems that you don't have access to internet.
18:52 scolby sbog_:  It seems that way, however on the Master node if I do "dig archive.ubuntu.com" it resolves the FQDN just fine.
18:52 sbog_ hmm. Where did you get such error?
18:54 scolby sbog_: the text of the error shown above appears on the Fuel GUI on the 'Dashboard' tab.  If I go to the 'Networks' tab and click the Verify Networks button I get the same error.
18:58 sbog_ Recheck your network settings, and if it looks okay - it sounds like a bug for me. Could you then file it, please?
19:11 mwhahaha i assume network checker is what is throwing that error
19:11 mwhahaha so are you sure that your nodes to be deployed to have access to the internet given the network configuration you have created for the environment?
19:12 sbog_ Seems that yes, but it should't popup such error if network settings is okay. So it must be checked first.
19:12 mwhahaha not that the settings are invalid but if the ip/vlans are wrong it would
19:12 mwhahaha ie switch misconfiguration
19:14 mwhahaha where the public network doesn't actually have internets
19:14 mwhahaha or something of the sort
19:29 scolby mwhahaha:  I'm running the "fuel-createmirror" script now on the Master - that seems to be working fine.  The idea is to make the slave nodes talk only to the Master node instead of the internet.
19:30 mwhahaha right that'll work but you may need to manually remove the internet repos from your environment after running fuel-createmirror
19:31 scolby mwhahaha:  Check, will do.  This is just a VirtualBox 'all-in-one' 7.0 install so I have no problem with savaging it.
19:32 sergmelikyan joined #fuel
19:40 bpiotrowski mwhahaha: are you going to run for fuellib lead?
19:41 mwhahaha not this time
19:43 bpiotrowski http://shonaliburke.com/wp-content/uploads/2011/06/4701379139_04f9905bd2.jpg
19:48 ppradhan hello all, my deployment failed with "error" on 3 nodes out of 10 nodes. I problem was incorrect network port setting. How can I deploy again without doing the whole thing once again?
19:48 mwhahaha you should be able to just run the deploy on the environment
19:49 mwhahaha it should target the failed nodes
19:49 ppradhan ok let me try
19:50 bpiotrowski if webui doesn't allow to do that, fuelclient certainly does
19:51 sbog__ joined #fuel
19:53 angdraug joined #fuel
20:21 ppradhan the redploy works
20:24 ppradhan I have 3 compute nodes . And I can see only one when doing nova-manage service list from the controller node.
20:24 ppradhan restart nova-compute on comptes
20:24 ppradhan ?
20:25 vsedelnik joined #fuel
20:35 subscope joined #fuel
20:39 Zer0Byte__ joined #fuel
20:57 Zer0Byte__ joined #fuel
21:05 Sesso_ are the 3 that failed in bootstrap?
21:14 ppradhan Sesso_: it came back after a while.
21:15 Sesso_ it takes a few minutes :)
21:19 Zer0BYte__ joined #fuel
21:24 az__ joined #fuel
21:36 RageLtMan joined #fuel
21:36 RageLtMan I've installed Fuel 7 GA and deployed a new openstack from it, but i apparently still cant upload ISO images - swift log says "Keystone reply error: status=404 reason=Not Found"
21:37 RageLtMan Fresh install, using the kibana, lma, vpnaas, fwaas plugins built from stackforge GH repo on Thursday (master branches). Seems to happen with or without the plugins though
21:37 gerst sounds like your install has failed
21:37 RageLtMan install completes
21:37 RageLtMan no errors logged
21:37 gerst I mean, do you use the new vxlan tunneling?
21:38 RageLtMan yes
21:38 gerst did you raise mtu for the private networks?
21:38 RageLtMan the private nets run on Mellanox adapters and switch which seem to stop working altogether when i screw with the MTU
21:39 RageLtMan Have a quick link to the doc which describes the procedure?
21:39 RageLtMan (if not i'm looking anyway, thanks though)
21:40 gerst the install we did also uses vxlan and it works, but we raised mtu to 1564 I think
21:41 gerst only thing it gets messy with if you share the private network interface with other interfaces
21:42 gerst http://docs.openstack.org/juno/config-reference/content/networking-options-plugins-ml2.html
21:43 Zer0Byte__ joined #fuel
21:44 gerst in Fuel its the interface assign window where you assign interfaces for networks and you set private network's interfaces mtu to 1550, althought if you dont' have a cisco switch I think it was supposed to be 1564
21:57 sbog_ joined #fuel
22:02 RageLtMan gerst: this problem isnt occuring each time, and while i'm a bit loathe to screw with the NICs MTU, i did set the switch ports to maximum
22:04 RageLtMan Also, this - http://docs.openstack.org/icehouse/install-guide/install/yum/content/neutron-ml2-network-node.html, talks about configuring MTU via DHCP, but i'm a bit unsure how to get neutron to use the expanded NIC MTU after install. Suppose i could just reprovision
22:11 Zer0Byte__ joined #fuel
22:22 justif joined #fuel
22:25 skath joined #fuel
22:27 kozhukalov_ joined #fuel
22:34 teran joined #fuel
22:34 alex_didenko joined #fuel
23:11 maximov joined #fuel
23:15 kate_ joined #fuel
23:15 zhangjn joined #fuel
23:23 kozhukalov_ joined #fuel
23:28 pumphouse joined #fuel
23:50 zhangjn joined #fuel

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