Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2014-03-27

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

All times shown according to UTC.

Time Nick Message
01:15 IlyaE joined #fuel
01:24 xarses joined #fuel
01:34 dubmusic joined #fuel
01:36 IlyaE joined #fuel
01:49 DaveJ__ joined #fuel
02:05 justif joined #fuel
02:07 fandi joined #fuel
02:30 vkozhukalov joined #fuel
03:07 ToTsiroll joined #fuel
03:32 fandikurnia joined #fuel
03:54 fandikurnia joined #fuel
03:54 fandi left #fuel
04:38 saju_m joined #fuel
05:22 dburmistrov joined #fuel
05:26 Ch00k joined #fuel
05:35 dburmistrov joined #fuel
06:23 IlyaE joined #fuel
06:49 Ch00k joined #fuel
06:50 saju_m joined #fuel
06:56 larion89 joined #fuel
06:57 larion89 i wrote and asked about using fuel in virtual environment. I still have some problems even tho i have opened up my vswitches with promiscious mode, i get this as one of the errors
06:57 larion89 (/Stage[main]/Ceph::Conf/Exec[ceph-deploy config pull]/returns) change from notrun to 0 failed: ceph-deploy --overwrite-conf config pull node-14 returned 1 instead of one of [0] at /etc/puppet/modules/ceph/manifests/conf.pp:44
06:59 larion89 if someone could help i would appreciate  it :)
07:38 vkozhukalov joined #fuel
07:39 AndreyDanin joined #fuel
07:54 AndreyDanin joined #fuel
08:02 larion89 is none capable of helping out? :)
08:06 Ch00k joined #fuel
08:09 meow-nofer_ joined #fuel
08:14 AndreyDanin joined #fuel
08:14 meow-nofer was kicked by ogelbukh: meow-nofer
08:15 meow-nofer_ was kicked by ogelbukh: meow-nofer_
08:24 evg_ joined #fuel
08:24 AndreyDanin joined #fuel
08:34 vk joined #fuel
08:37 topochan joined #fuel
08:50 Ch00k_ joined #fuel
08:53 meow-nofer_ joined #fuel
08:55 saju_m joined #fuel
08:58 meow-nofer_ joined #fuel
08:59 Ch00k joined #fuel
09:09 rvyalov joined #fuel
09:20 Ch00k joined #fuel
09:28 saju_m joined #fuel
09:33 e0ne joined #fuel
09:41 anotchenko joined #fuel
09:47 mirrorbox joined #fuel
10:00 evg larion89, could you, pls, describe your networking. Are you sure you have full connectivity in adm & storage networks?
10:01 e0ne_ joined #fuel
10:02 larion89 i think i've solved it actually :) i just repeated the process with installing fuel once again and then i deployed it node by node, so probably it was something with some differences in the configurations of the virtual machines
10:03 larion89 i'm installing the storage node now, i've got one controller and a compute node installed and done.
10:03 larion89 i'm gonna try to install this on baremetal machines later today IF i have enough machines to deploy on
10:05 larion89 btw, before i had everything connected to one vswitch, now i separated everything and turned promiscious mode on, and it went well. so i guess it was something with that, that caused the problem
10:25 larion89 evg, i cant install the storage node, i get some weird errors
10:25 larion89 2014-03-27 10:22:29WARNING [1633] Disabling the upload of disk image because deploy ended with an error
10:25 larion89 what might cause this?
10:44 evg larion89, it's a warning, but what was the error (above in the log)?
10:54 evg larion89, and i still doubt about your network configuration. is it really necessary for you to run promiscuous mode.
11:09 evg_ joined #fuel
11:11 Ch00k joined #fuel
11:14 saju_m joined #fuel
11:23 fandikurnia joined #fuel
11:32 mihgen joined #fuel
11:35 saju_m joined #fuel
11:56 larion89 evg, i dont know, might be something weird with my network configuration. Though i get some other weird things aswell. Even tho my compute node is "ready" i cant see it in my dashboard....I'm a very big newbie when it comes to this environment
12:00 justif joined #fuel
12:04 anotchenko joined #fuel
12:36 evg larion89, ready in fuel? and don't see them in "dashboard->admin->system info->compute services"?
12:37 larion89 no, under hypervisor :/
12:37 larion89 it's up in system info
12:39 larion89 but there's nothing under hypervisor. my healthcheck also fails. everything, including storage is ready now
12:39 larion89 but it still doesnt work
12:42 mihgen larion89: so what Network Verify says to you?
12:42 mihgen and what if you do simple "nova-manage service list" on one of your openstack nodes?
12:42 larion89 Verification failed. Network verification on Neutron is not implemented yet
12:43 larion89 even tho it's implemented
12:43 larion89 i can see router and all that in the dashboard
12:43 larion89 so that means that the communication between my controller and the compute/storage doesnt work?
12:45 dubmusic joined #fuel
12:46 larion89 i've got separate storage and compute nodes ofc, if it was unclear
12:46 justif what OS did you delploy with?
12:46 Dr_Drache joined #fuel
12:46 larion89 ubuntu 12.04, the one that is coming with the newest fuelversion
12:47 larion89 i tried centOS earlier on, but it failed aswell
12:48 larion89 or what did you mean justif
12:48 larion89 ?
12:48 mihgen oh yep … (
12:48 mihgen you had to run network verification before deployment
12:48 mihgen it would show you if you have all required VLAN passing
12:49 mihgen now it's impossible to check, as network checker works only with Linux VLANs
12:49 mihgen and doesn't work with openvswitch
12:49 mihgen it's likely that you have issues with management network.
12:49 mihgen so what about "nova-manage service list" ?
12:50 larion89 so what do you suggest? i deploy without using "SDN"?
12:50 larion89 as in flat networks?
12:50 larion89 or what it's called :p
12:50 larion89 sec mihgen!
12:51 mihgen larion89: I suggest first of all check network manually now
12:51 MiroslavAnashkin larion89: Have you read this guide before configuring your networks? All these settings, including NIC model works for ESXi as well.  http://docs.mirantis.com/fuel/fuel-4.1/install-guide.html#virtualbox
12:51 justif well centos needs the OVS VLAN splinters option due to the older kernel, Ubuntu should work without issue
12:52 mihgen justif: oh yep centos amazes :)
12:55 larion89 i dont even get a freaking "Login" now. Gotta start all over again i believe. i get "end_request: I/O error, dev fd0 sector 0". dont know what i've done :p
12:56 larion89 i'll check that link
12:59 e0ne joined #fuel
13:04 larion89 basically what i've done when it comes to network is, i've got 4 networkcards for each node. first one is for pxenetwork(fuel) the 2nd one is for internal, the 3rd one is for external and 4th is for floating IP. isnt that correct? I've got promiscious mode on "accept" on ALL these switches.
13:07 e0ne joined #fuel
13:10 larion89 and this is the vswitches i'm talking about, the names.
13:13 larion89 i'll start all over again then :)
13:15 MiroslavAnashkin the 3rd one is for external and 4th is for floating IP. - Floating IP range is sub-range from external network
13:15 MiroslavAnashkin Both should live on the same network interface
13:20 larion89 ah okey! i'll fix that
13:22 dburmistrov joined #fuel
13:25 dubmusic joined #fuel
13:27 Ch00k joined #fuel
13:30 brain461 joined #fuel
13:48 meow-nofer_ joined #fuel
13:48 MiroslavAnashkin joined #fuel
13:48 evgeniyl` joined #fuel
14:01 evg joined #fuel
14:05 dubmusic larion89.  Your NIC layout does not look correct.  Also, the Network check with VLAN config never works with the CentOS bootstrap OS doing the checking (driver/kernel issue).  ONce you are in Ubuntu, you can manually check your connectivity over the VLANs
14:11 larion89 i just removed the vlan tagging now, and added more cards instead. i also went for ubuntu and nova network. the verify network came through aswell, so i suppose it's going to work now.
14:11 larion89 it's installing as we speak
14:11 rmoe joined #fuel
14:17 alexdepalex joined #fuel
14:33 mihgen joined #fuel
14:43 alex_didenko joined #fuel
14:43 jobewan joined #fuel
14:46 larion89 Explaint to me why i get errors which doesnt even get into the logs? :S
14:47 larion89 Deployment has failed. Check these nodes: 'Untitled (a0:45)'
14:47 larion89 if i take a look at that server in the fuel logs
14:48 larion89 i dont get anything that has gone wrong
14:48 larion89 no errors no warning
14:49 dubmusic Which logs are you looking at and where? (/var/log/remote on Fuel-master) or GUI?
14:49 larion89 GUI
14:52 jobewan joined #fuel
14:52 larion89 i tried to deploy controller+compute+storage at the same time. i guess i do it once again step by step again then.
14:54 Axam joined #fuel
14:56 dubmusic Is this HA?
14:56 larion89 no HA
14:56 Dr_Drache vlans are, difficult at best with fuel
14:58 IlyaE joined #fuel
14:59 larion89 now i used Ubuntu. and i tried to deploy everything at the same time, i get error in deployment but no errors in puppet telling me that there was any errors on the node i was getting errors from
15:00 Dr_Drache larion89, what was the error agian?
15:00 larion89 there's none
15:00 dubmusic I actually have VLANs all working, nicely, but there were many issues with Fuel that threw me off course.
15:00 larion89 Deployment has failed. Check these nodes: 'Untitled (a0:45)' just this
15:01 larion89 but
15:01 larion89 if i take a look in the gui log
15:01 larion89 there's no errors what so ever from that node.
15:01 dubmusic which version of Fuel?
15:01 larion89 4.1
15:01 larion89 i downloaded it yesterday
15:01 dubmusic OK
15:02 dubmusic Does Fuel get past the bootstrap phase?
15:02 angdraug joined #fuel
15:03 Dr_Drache larion there has to be an error somewhere. it may not make sense, but there is one
15:03 larion89 well it make sense that there's any errors
15:03 larion89 :p
15:04 larion89 there's none from this deployment tries
15:04 dubmusic If there is nothing in the Puppet logs, it may not be getting to that phase.  Does it finish installing Ubuntu on all the nodes.
15:05 larion89 yes
15:05 larion89 or else the isntall openstack doesnt start
15:05 larion89 i assume?
15:05 larion89 the other nodes are "ready"
15:05 larion89 <<
15:05 larion89 :<<
15:05 dubmusic yep.  I am just trying understand where you are in the install
15:05 Dr_Drache debugging on?>
15:06 larion89 not what i know if
15:06 larion89 of*
15:06 dubmusic How many nodes?
15:06 larion89 how do i turn that on?
15:06 Dr_Drache larion89, in settings, before you deployed.
15:07 larion89 alright
15:07 larion89 i'll do like this again then. i'll reset
15:07 larion89 and start debug thing. and deploy again
15:07 larion89 and we'll see how it goes :p
15:07 larion89 10th try! :p
15:07 larion89 my face is literally looking liek this >.<
15:10 dubmusic How many NICs, Speed, Vendor?
15:11 larion89 it's one esxi host. and virtual machines
15:11 larion89 and i have promisc mode on
15:11 larion89 i have one nic for each thing. including vswitches
15:11 larion89 none of them are connected to internet, i assume that shouldnt make any differences
15:11 larion89 and the verify network gave me green light
15:12 larion89 so i dont understand
15:12 Dr_Drache verify doesn't work 100% in neutron.
15:12 Dr_Drache not for VLANs
15:12 larion89 not using vlans
15:13 Dr_Drache neutron GRE untagged?
15:13 larion89 no nova network. the one furtherst up
15:14 Dr_Drache is there a need for nova, or not?
15:15 larion89 nova isnt that the compute node?
15:15 larion89 when you add the openstack environment, i took the standard one on networking.
15:16 larion89 are you with me? :)
15:18 e0ne joined #fuel
15:20 Dr_Drache yea
15:20 Dr_Drache neutron is recommended unless you need nova-networking
15:22 larion89 i'll try that after i've tried this..... :D
15:22 larion89 i know that SDN is ofc recommended :p
15:23 IlyaE joined #fuel
15:26 MiroslavAnashkin larion89: Please check the following logs: 1. In Fuel UI select your failed environment, select Logs tab, select Other Servers. select your failed node. select Err as Min.level and switch between sources clisking Show button
15:28 Ch00k joined #fuel
15:28 larion89 i did EXACTLY that. no errors
15:29 larion89 except some errors on bootstrap but that was nothing that was involved by this deployment try.
15:29 Dr_Drache larion89, they are all SDN
15:29 Dr_Drache just one is new one is old/being removed.
15:30 larion89 ah okey. :p
15:30 Dr_Drache in general, if you don't have a need that neutron doesn't support, use neutron.
15:35 Ch00k joined #fuel
15:43 larion89 so you guys have any tips of what to do while waiting? :P
15:43 larion89 51% so far -.-
15:44 Dr_Drache lol
15:50 larion89 now the controller came through
15:50 larion89 installing compute and storage now
15:50 larion89 we'll see how it goes
15:51 larion89 and i changed nothing at all. which is a bit funny.
15:53 rustyrobot joined #fuel
15:56 mihgen joined #fuel
16:00 vk joined #fuel
16:05 larion89 okey
16:06 larion89 now everything went through clean and nice. BUT i cant get into my dashboard!
16:07 larion89 i get something went wrong thingy
16:07 Dr_Drache why can't you?
16:07 Dr_Drache routes maybe?
16:07 Dr_Drache try both ips?
16:07 larion89 havent done that gonna do it
16:10 MiroslavAnashkin larion89: It is definitely route issue. What is your host OS?
16:11 larion89 Host OS? for the controller?
16:11 larion89 ubuntu this time
16:12 larion89 if it has contact and somethign went wrong. it cant be that it's put on the wrong switch either?
16:12 MiroslavAnashkin Ah, you use ESXi.
16:12 larion89 yes
16:15 MiroslavAnashkin And your OpenStack dashboard is not accessible by default 172.16.0.XX address?
16:16 larion89 i'm using the 10.20.0.10 and get "something went wrong contant adminsitrator" and so on
16:16 larion89 so there's some contact.
16:17 larion89 which is weird. it has been working before.
16:17 larion89 everythign else has nto been working, but the controller i havent had any issues with before
16:19 larion89 gonna try to add another card to this machine and move it and see if there's some difference as i know management is still on one of the vswitches.
16:19 larion89 i dont know which one. but it's weird taht i get connected anyhow
16:19 MiroslavAnashkin Please check what is default virtual IP address on your controller  (by default 172.16.0.) and try to connect through it
16:19 larion89 and i can ping as it is right now
16:22 larion89 is it going under management or admin(PXE)?
16:23 larion89 i tried that earlier MiroslavAnashkin
16:23 larion89 changing the IP on this machine, on my 2nd card
16:24 Ch00k joined #fuel
16:24 evg joined #fuel
16:32 larion89 Deployment of environment 'test' is done. Access the OpenStack dashboard (Horizon) at http://172.16.0.2/ or via internal network at http://10.20.0.10/
16:32 larion89 this is what it says
16:33 larion89 GAH!
16:33 larion89 i dont get it
16:34 MiroslavAnashkin ?
16:34 larion89 i cant administrate it
16:34 larion89 now when EVERYTHING else works
16:34 larion89 the dashboard is not working
16:34 larion89 can i restart it?
16:35 larion89 as i dont have a login prompt on my machines
16:35 larion89 i cant do anything without restarting them either
16:35 larion89 i dont dare to
16:35 larion89 but i guess i'm forced to.
16:36 larion89 why isnt there restart buttons? :p
16:42 larion89 the dashboard is dead
16:43 larion89 i restarted all nodes
16:43 larion89 same thing
16:52 dburmistrov joined #fuel
16:54 mattymo joined #fuel
16:56 MiroslavAnashkin No, it is not nodes. What does `arping <one of installed controllers IP> says?
16:56 MiroslavAnashkin Please run arping under root
16:58 rmoe joined #fuel
17:00 evg
17:02 larion89 help me here. what's the name of the dashboard?
17:08 MiroslavAnashkin Dashboard is on the default port. so you only have to know one of machine IP addresses to reach it
17:12 Ch00k joined #fuel
17:18 e0ne joined #fuel
17:33 e0ne joined #fuel
17:38 xarses joined #fuel
18:00 e0ne joined #fuel
18:12 dubmusic joined #fuel
18:16 angdraug joined #fuel
18:26 Ch00k joined #fuel
18:31 fandi joined #fuel
18:33 Ch00k joined #fuel
18:45 larion89 ITS WORKING! fucking finally. i solved it by removing some old cookies and stuff..........
18:45 larion89 sigh.
18:47 MiroslavAnashkin Grats!
18:51 pet159 joined #fuel
18:53 pet159 I have the Fuel master node deployed on a bare metal machine. I tried to deploy some nodes. They are recognized but the deployment never finishes when I create a control and compute node. When I look at the machines they are stuck at the "bootstrap login". I try giving it a login but it does not help.
19:05 MiroslavAnashkin pet159: Have you created new environment in Fuel master node UI and added your discovered nodes as controllers, computes, etc?
19:11 MiroslavAnashkin Have you clicked Deploy Changes button?
19:26 Ch00k joined #fuel
19:31 vkozhukalov joined #fuel
19:47 IlyaE joined #fuel
20:03 thehybridtech joined #fuel
21:06 xarses joined #fuel
21:16 rmoe_ joined #fuel
21:32 angdraug joined #fuel
21:43 tatyana joined #fuel
22:00 tatyana left #fuel
22:00 tatyana joined #fuel
22:11 AndreyDanin joined #fuel
22:20 rmoe joined #fuel
22:23 xarses joined #fuel
22:25 tatyana left #fuel
23:19 justif Not sure if a bug or not but on the first deploy the private net_04 network and router_04 do not work.
23:19 justif I have do delete them and re add then everything is kosher
23:21 xarses possibly a bug, but it should fail our ci tests also
23:22 xarses does this happen all the time?
23:43 anotchenko joined #fuel

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