Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2014-04-03

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

All times shown according to UTC.

Time Nick Message
01:04 rmoe joined #fuel
01:19 piersdd joined #fuel
02:49 thehybridtech joined #fuel
03:52 dburmistrov joined #fuel
04:37 mihgen joined #fuel
04:56 xarses joined #fuel
05:00 piersdd joined #fuel
05:05 Ch00k joined #fuel
05:15 dburmistrov joined #fuel
05:30 IlyaE joined #fuel
06:16 Ch00k joined #fuel
06:21 alex_didenko joined #fuel
06:30 al_ex joined #fuel
06:52 AndreyDanin_ joined #fuel
06:54 e0ne joined #fuel
07:20 sb joined #fuel
07:20 strictlyb joined #fuel
07:25 e0ne_ joined #fuel
07:27 akislitsky joined #fuel
07:51 Ch00k joined #fuel
07:55 topochan joined #fuel
08:34 mihgen joined #fuel
08:55 e0ne joined #fuel
09:08 e0ne_ joined #fuel
09:19 neophy joined #fuel
09:23 neophy I am new to fuel. Just installed ISO in one Physical Machine. I am not able to access web UI through port 8000 or not able to ssh to fuel machine. Any clue will help to troubleshoot the issue.
09:23 neophy fuel is installed in remote machine
09:38 aglarendil|nb joined #fuel
09:45 neophy any help?
09:47 mattymo neophy, are you able to access http://(ip of your host):8080 ?
09:47 mattymo if you look at the shell on your physical system, it shows the IP address it uses
09:47 mattymo the default address is 10.20.0.2 (which is probably not the network you use)
09:49 neophy I have configured ip address at the time of installation
09:49 neophy I am not able to access the url
09:50 neophy I am table to telnet (ipaddress):8000 from same machine
09:51 neophy but not able to access from my client machine browser
09:52 neophy mattymo: Is there any firewall rule I have enable in fuel to access from client machine browser
09:52 neophy ?
09:53 aglarendil|nb neophy: do you have a vnc or IPMI console to your master node?
09:53 neophy aglarendil|nb: I have iDRAC console
09:54 mattymo neophy, you need to read the Fuel documentation first for networking considerations
09:54 mattymo then you should reinstall the host and enable the Fuel menu on next boot (also in the docs)
09:54 aglarendil|nb neophy: can you login to the  node and issue telnet from the local host
09:54 aglarendil|nb neophy: can you login to the  node and issue telnet from the local host?
09:54 mattymo http://docs.mirantis.com/fuel/fuel-4.1/
09:56 neophy aglarendil|nb: Yes. I loged into master node through iLO console. I am able to telnet from localhost
09:57 aglarendil|nb neophy: can you issue "GET /" request on the http://<master_ip>:8000 from the iLO console ?
10:07 neophy aglarendil|nb: It returns html page. But I am not able to copy it to show you
10:08 aglarendil|nb is it http 200 ?
10:08 aglarendil|nb I mean status code
10:09 neophy no status code 200
10:17 neophy aglarendil|nb: $curl "http://<master_ip>:8000" return http://derp.co.uk/67c5c
10:18 aglarendil|nb can you issue curl -v ?
10:20 Akshik joined #fuel
10:22 DaveJ__ joined #fuel
10:28 mattymo neophy, the Fuel application works fine. You can only configure networking during installation, though
10:29 mattymo aglarendil, the curl works. it's as expected. his networking isn't configured and going through any troubleshooting of curl isn't going to get him closer to being able to test out Fuel
10:29 neophy aglarendil|nb:  I ran curl -v <master_node>:8000 > file ; file data screen  : http://derp.co.uk/6689f and  curl status screen : http://derp.co.uk/9734d .
10:29 mattymo aglarendil|nb, &
10:29 mattymo aglarendil|nb, ^
10:33 neophy mattymo: I have to reinstall Fuel. Right?
10:33 mattymo yes
10:34 mattymo neophy, read the pre-installation documentation first. you should have 1 interface pointing to the external network and 1 interface that will be managed by Fuel (where it can run DHCP for deployment)
10:38 neophy mattymo: ok. I will read pre-installation documentation then reinstall Fuel again. Is it show fuelmenu to do network configuration post installation right?
10:38 mattymo yes
10:53 neophy mattymo:  Next installation I will use eth0 for Network Setup and eth1 for PXE Setup. Do you have some doc which will explain about fuelmenu configuration? I am new to PXE boot that is why I not able to get clear picture about static pool and DHCP pool
10:55 neophy I installed it two times and did the same network configuration mistake I guss
11:03 evg joined #fuel
11:04 warpig hi guys, was wondering if there is an update on disk detection with ubuntu on Fuel 4.1...
11:05 warpig I've upgraded to 4.1.1, but the ubuntu PXE image doesn't see any disks
11:09 e0ne joined #fuel
11:11 e0ne_ joined #fuel
11:11 mattymo neophy, try it out and let me know if you get stuck
11:12 mattymo I tried to make it pretty obvious
11:13 e0ne__ joined #fuel
11:13 saibarspeis Does anybody know a good reason why openstack dashboard won't work if you deploy several clusters with the same FUEL server ?
11:14 saibarspeis I have reached the end of the trials with FUEL and OpenStack, the first environment works fine, it apears operational in Fuel and I can access it's dashboard
11:14 saibarspeis The second and third environments also are Operational in Fuel but their dashboard won't work
11:14 saibarspeis I even installed an apache with mod_proxy to break the protocol and just be inside the networks in which FUEL and the nodes operate
11:15 saibarspeis Using the apache with mod_proxy the first cluster works but the second and third don't
11:15 mattymo saibarspeis, sounds like you hit a bug. I'm not familiar with it, but could you possibly file a bug with data on bugs.launchpad.net/fuel?
11:15 saibarspeis Is there a limitation ? Something that will prevent OpenStack dashboard to listen to the ports ?
11:16 warpig saibarspeis: I assume you specified a different subnet for each env?
11:16 saibarspeis Yes
11:16 saibarspeis I've broken apart a /24
11:16 anotchenko joined #fuel
11:16 saibarspeis Into four /26 networks
11:16 saibarspeis For all the interfaces
11:16 saibarspeis That is admin, public, storage,management
11:16 saibarspeis The only common ground is the PXE interface for these clusters
11:16 saibarspeis And my IP allocations work because I can see the servers from outside
11:17 saibarspeis Ahh and Network verification in my case is succeding each time
11:17 saibarspeis (I have redeployed these three clusters 7 times now)
11:17 saibarspeis I know the configurations by hard :)
11:18 mihgen > but their dashboard won't work - can you be more precise, do you able to ping host?
11:18 mihgen telnet to the port?
11:18 mihgen over what network you are trying to access it, public or admin PXE ?
11:18 saibarspeis telnet works
11:19 saibarspeis GET / it says that I need to be redirected to http://192.168.10.2/horizon/
11:19 saibarspeis For all environments this response is present if I do GET / on 80
11:19 saibarspeis But in the browser it won't work, it will just timeout
11:19 saibarspeis Only the first cluster/instance works fine
11:20 saibarspeis And each time i redeploy this happens
11:20 saibarspeis Only DEV-ENV01 works
11:20 mihgen do you have HA installation for 2nd & 3rd envs?
11:20 saibarspeis DEV-ENV02 and DEV-ENV03 timeouts and I can see the dashboard for the first 5 minutes, then it goes away
11:20 saibarspeis NO
11:20 saibarspeis My configuration is like this
11:21 saibarspeis Ubuntu Ubuntu + Ceilometer
11:21 saibarspeis QEMU
11:21 saibarspeis Cinder LVM for ISCSI
11:21 saibarspeis Neutron with GRE segmentation
11:21 saibarspeis (because I actually use the same L2 segments)
11:21 saibarspeis Verify network is successful for all the three environments
11:22 saibarspeis The deployment is completed all the time with no errors
11:22 saibarspeis All the env show operational status
11:22 saibarspeis (I can run healthchecks on them)
11:22 mihgen do you run telnet from your host?
11:22 saibarspeis From my host I cannot run directly on their 80 port
11:22 saibarspeis I am nated
11:23 saibarspeis Or apache mod_proxy (I have devised two ways because actually the dashboard is on two interfaces)
11:23 saibarspeis On the NAT (public) the dashboard times out for the DEV02 and DEV03 and works fine for DEV01
11:23 saibarspeis On the apache_proxy (admin interface) is the same
11:23 saibarspeis It will timeout and my proxy will respond that the remote server could not handle the request
11:23 saibarspeis And DEV01 actually works fine
11:24 mihgen so can you try to tcpdump the traffic and actually see incoming requests from telnet and browser to see the difference?
11:24 saibarspeis This apache proxy is a linux (rhel based) inside the admin network of the servers
11:24 mihgen if you say that telnet works fine, but not browser
11:24 saibarspeis Yes I can see them all
11:25 saibarspeis I have traced them to br-mgmt
11:25 saibarspeis And on br-mgmt a vhost is listening
11:25 saibarspeis But it's on port 8888 ?
11:26 saibarspeis In api_proxy.conf for apache
11:26 vk joined #fuel
11:27 saibarspeis There is a Proxy directive that denies everything except for the FUEL server
11:27 saibarspeis In no case the requests come from the fuel server
11:27 saibarspeis Or that is only for the internal API .. i get it now
11:27 saibarspeis It's for fuel to talk to the API of the openstack
11:28 saibarspeis Ok so port 80 is the default vhost
11:28 saibarspeis Which redirects to the /horizon folder
11:29 saibarspeis Which is actually served from /usr/share/openstack-dashboard/openstack_dashboard/wsgi
11:29 saibarspeis And is allowed from all
11:29 saibarspeis I don't know why only on DEV01 would work
11:30 mihgen what do you see in apache/horizon logs?
11:30 saibarspeis Nothing
11:31 saibarspeis That's why I'm trying to see with tcpdump
11:31 saibarspeis br-fw-admin is the interface which should hold the admin ip
11:31 saibarspeis And that checks out
11:33 saibarspeis The requests are comming from the correct source on the correct interface
11:34 saibarspeis I can paste the tcpdump output here but I'll probably get banned
11:35 saibarspeis http://pastebin.com/UFxcsDpX
11:35 saibarspeis Proxy Error
11:35 saibarspeis The proxy server received an invalid response from an upstream server.
11:35 saibarspeis The proxy server could not handle the request GET /horizon/admin/.
11:35 saibarspeis Reason: Error reading from remote server
11:35 saibarspeis Apache/2.2.15 (Red Hat) Server at 172.16.240.158 Port 8002
11:36 saibarspeis On 8001 (DEV01 env) the same request works
11:36 saibarspeis There are no lines in any apache logs
11:36 mihgen can you access horizon directly?
11:37 mihgen is it centos install?
11:37 saibarspeis No, it's ubuntu
11:37 saibarspeis And no, I can't access it directly, I have no windows machines in that environment
11:37 saibarspeis But I can show you this from the proxy server
11:37 mihgen can you try ssh tunnel then to master node then
11:38 saibarspeis Yes I can try
11:38 mihgen ssh -2NfL 8000:node-1:80 root@10.20.0.2
11:38 mihgen something like this
11:38 saibarspeis Just a second to give you another pastebin
11:38 mihgen where node-1 is your controller
11:39 AndreyDanin_ joined #fuel
11:40 saibarspeis http://pastebin.com/2dqa25Xv
11:40 saibarspeis And now ssh tunnel
11:42 e0ne joined #fuel
11:44 saibarspeis It's working with ssh tunnel
11:45 saibarspeis 192.168.10.66:80 192.168.194.254 - - [03/Apr/2014:11:44:01 +0000] "GET /horizon/admin/defaults/ HTTP/1.1" 200 4785 "http://localhost:40008/horizon/admin/projects/" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:28.0) Gecko/20100101 Firefox/28.0"
11:45 saibarspeis 192.168.10.66:80 192.168.194.254 - - [03/Apr/2014:11:44:02 +0000] "GET /horizon/i18n/js/horizon/ HTTP/1.1" 200 4145 "http://localhost:40008/horizon/admin/defaults/" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:28.0) Gecko/20100101 Firefox/28.0"
11:45 saibarspeis 192.168.10.66:80 192.168.194.254 - - [03/Apr/2014:11:44:01 +0000] "GET /horizon/admin/defaults/ HTTP/1.1" 200 4785 "http://localhost:40008/horizon/admin/projects/" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:28.0) Gecko/20100101 Firefox/28.0"
11:45 saibarspeis 192.168.10.66:80 192.168.194.254 - - [03/Apr/2014:11:44:02 +0000] "GET /horizon/i18n/js/horizon/ HTTP/1.1" 200 4145 "http://localhost:40008/horizon/admin/defaults/" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:28.0) Gecko/20100101 Firefox/28.0"
11:45 saibarspeis I have made the tunnel using the same server which is setup as a proxy server for the dashboards
11:46 mihgen looks like your proxy is not configured properly?
11:46 saibarspeis Looks like it
11:46 saibarspeis But works for the first ENV and not for the next two ?
11:46 mihgen may be you forget to restart apache after changing config...?
11:46 saibarspeis And in pasebin I showed you the configurations are actually the same, they have been copy/pasted and only the last bit is changed
11:47 saibarspeis :))
11:47 saibarspeis Hahha
11:47 saibarspeis No
11:47 saibarspeis [root@openw-rebound ~]# httpd -S
11:47 saibarspeis VirtualHost configuration:
11:47 saibarspeis wildcard NameVirtualHosts and _default_ servers:
11:47 saibarspeis *:8003                 is a NameVirtualHost
11:47 saibarspeis default server 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:17)
11:47 saibarspeis port 8003 namevhost 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:17)
11:47 saibarspeis *:8002                 is a NameVirtualHost
11:47 saibarspeis default server 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:9)
11:47 saibarspeis port 8002 namevhost 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:9)
11:47 saibarspeis *:8001                 is a NameVirtualHost
11:47 saibarspeis default server 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:1)
11:47 saibarspeis port 8001 namevhost 172.16.240.158 (/etc/httpd/conf.d/openstack.conf:1)
11:47 saibarspeis Syntax OK
11:47 saibarspeis And this is the log
11:47 saibarspeis [Thu Apr 03 13:33:35 2014] [error] [client 172.16.198.118] (70007)The timeout specified has expired: proxy: error reading status line from remote server 192.168.194.5
11:47 saibarspeis [Thu Apr 03 13:33:35 2014] [error] [client 172.16.198.118] proxy: Error reading from remote server returned by /horizon/admin/
11:48 saibarspeis [Thu Apr 03 13:36:29 2014] [error] [client 172.16.198.118] (70007)The timeout specified has expired: proxy: error reading status line from remote server 192.168.194.5
11:48 saibarspeis [Thu Apr 03 13:36:29 2014] [error] [client 172.16.198.118] proxy: Error reading from remote server returned by /horizon/admin/
11:48 mihgen please use paste services..
11:48 saibarspeis ok
11:48 saibarspeis There's nothing more to show anyway
11:48 mihgen don't know, looks strange..
11:53 saibarspeis I might deliver this openstack environment like this, and tell the devs to use SSH tunnels
11:53 saibarspeis And it might be ok
11:53 saibarspeis But it should not do this
11:53 saibarspeis Please tell me if any of you guys deployed MULTIPLE clusters with the same fuel server, with GRE segmentation and it does the same thing with the dashboard
11:54 mihgen what do you mean - does the same thing?
11:54 mihgen it works fine actually if you have multiple envs..
11:54 saibarspeis I mean it doesen't allow you to access the dashboard using a DNAT or a proxy
11:54 saibarspeis Just directly connected TCP sessions
11:54 mihgen even your first env works over proxy
11:55 saibarspeis Well it does not look so fine to me here ... but I am new and I don't know all the detailes
11:55 mihgen it should be something with your proxy or some other networking ..
11:56 saibarspeis I don't think is that because the configuration is the same
11:56 saibarspeis And the connection comes the same way from the same network
11:56 saibarspeis And actually it works if I use that server as an SSH tunnel
11:56 saibarspeis It just doesen't like proxy or DNAT :)
11:57 mihgen what if you try to run squid proxy instead then?
12:01 saibarspeis Let me try that
12:01 saibarspeis It will take a little to install and configure it
12:20 dburmistrov_ joined #fuel
12:30 AndreyDanin__ joined #fuel
12:32 al_ex joined #fuel
12:47 justif joined #fuel
13:03 anotchenko joined #fuel
13:24 MiroslavAnashkin warpig: Have you selected Ubuntu from PXE boot menu? It is incorrect,  there is only CentOS bootstrap image
13:25 warpig no, I chose it as an environment OS
13:25 warpig (within Fuel)
13:26 warpig we spoke about this a while back and I haven't created a snapshot...  :S
13:27 MiroslavAnashkin Then, you may create and share diagnostic snapshot so I may look what is wrong with the disks. Does CentOS environment see these disks?
13:27 thehybridtech joined #fuel
13:27 warpig I'm generating one now...
13:27 warpig :D
13:28 warpig CentOS sees my disks.
13:29 warpig I compiled the hpsa kernel module for that kernel and added it to initrd.gz.  I can see it loading the module, but the disks still weren't seen.  I haven't checked the udev rules
13:32 warpig anywhere in particular to upload it?
13:36 BillTheKat joined #fuel
13:39 warpig bueller.... bueller....  ;)
13:44 aglarendil|nb joined #fuel
13:48 MiroslavAnashkin warpig: yes, dropbox, google drive - everyone usually works well
13:51 warpig ah, OK, cool...  gimme 2 min...
13:57 warpig MiroslavAnashkin: https://drive.google.com/file/d/0Bwjg9oB4ypV7Vm5odUQ0YVMzZlk/edit?usp=sharing
13:57 warpig MiroslavAnashkin: the environment consists of 4 HP ProLiant BL465c Gen8 with P220i controllers
13:58 warpig controller node in RAID1, compute nodes in RAID0
14:06 MiroslavAnashkin warpig: Please download and apply the latest Fuel 4.1 patch and check the issue still exists. http://download.mirantis.com/fuelweb/Fuel_updates/4.1/fuel_partition_manager_patch_41_to_411/fuel_partition_manager_patch_41_to_411_ver4.run
14:13 warpig MiroslavAnashkin: I've installed ver1 already.  I'll report back in a few min with the results...  Cheers!
14:38 saibarspeis Ok just so you know
14:38 saibarspeis Making a squid cache on the exact same server where I've used mod_proxy and creating vhosts for each port to point to each openstack dashboard ... WORKS
14:56 IlyaE joined #fuel
14:59 mihgen joined #fuel
15:01 warpig MiroslavAnashkin: still no love - same results, the installer starts and then when it's time to partition the disk, that screen just loops over and over
15:03 warpig MiroslavAnashkin: I have loaded a few screenshots of the problem if they help...
15:03 warpig https://drive.google.com/folderview?id=0Bwjg9oB4ypV7UU14Z0pWSG80NkE&amp;usp=sharing
15:04 warpig the hpsa driver is loaded
15:04 warpig fdisk doesn't see any disk
15:04 MiroslavAnashkin OK, I'll check
15:05 warpig dmesg reports:  hpsa 000:03:00.0 board not ready, timed out.
15:06 warpig I have installed Ubuntu 12.04.4 on that blade with a vanilla iso and the controller and disks were seen.
15:06 MiroslavAnashkin joined #fuel
15:24 al_ex2 joined #fuel
15:37 anotchenko joined #fuel
15:49 rmoe joined #fuel
16:07 mjeanson joined #fuel
16:26 jobewan joined #fuel
16:27 dburmistrov joined #fuel
16:27 mjeanson joined #fuel
16:34 MiroslavAnashkin warpig: Funny, but it looks like Ubuntu 12.04 has older hpsa driver, without support for HP P220i controllers
16:36 angdraug joined #fuel
16:39 xarses warpig: is this your issue? https://bugs.launchpad.net/fuel/+bug/1301704
16:50 e0ne joined #fuel
17:01 Ch00k joined #fuel
17:11 MiroslavAnashkin Yes, looks like this is it.
17:18 mutex so occasionally I am seeing a problem with the neutron-openvswitch-agent process on my compute nodes
17:19 mutex if I launch a new instance it simply will not be connected to the networking layer correctly
17:19 mutex the nova daemon receives the network connection messages, I see them in the debug log
17:19 mutex but the instance will be unable to get a DHCP address
17:19 mutex however, any instances that were already connected have no problem
17:20 mutex so it seems that it is a problem processing the rules/code to connect an instance to a network
17:20 mutex I briefly looked through the fuel bugs for openvswitch and did not see anything obvious
17:21 mutex simply restarting the openvswitch agent on the compute node fixes the problem
17:21 mutex but it would be nice if I didn't have to do this
17:42 Dr_Drache joined #fuel
17:42 Dr_Drache hey MiroslavAnashkin
17:42 Dr_Drache got a new one for you.
17:43 jobewan joined #fuel
17:44 MiroslavAnashkin ?
17:45 Dr_Drache cannot deploy with ver4 of the patch on 4.1
17:45 Dr_Drache I get constant timeout exceeded
17:46 Dr_Drache about 1/2 way through the compute's install of openstack
17:47 Dr_Drache snapshot is uploading
17:48 xarses joined #fuel
17:49 Dr_Drache MiroslavAnashkin, https://www.dropbox.com/s/fhl45tvvr1yj0ce/fuel-snapshot-2014-04-03_17-43-08.tgz
17:49 MiroslavAnashkin OMG, what so wrong might have happened with compute?
17:50 Dr_Drache I am not sure.
17:50 Dr_Drache I might unpatch ver4
17:50 Dr_Drache and go back.
17:50 Dr_Drache but it had a fix I wanted, so I applied it.
17:54 Dr_Drache MiroslavAnashkin, whatever causes it, was changed from V2 to V4
18:00 Dr_Drache MiroslavAnashkin, is there still a link for v2 avilble?
18:10 MiroslavAnashkin Hmm, no. We intentionally removed all previous versions.
18:10 jobewan joined #fuel
18:10 Dr_Drache I see that
18:12 MiroslavAnashkin Dr_Drache: Cannot find interface with ip which includes to admin subnet "Untitled (CE:7D). Please check if your network switch is up and running.
18:14 Dr_Drache wtf. thanks, I went from a running enviroment, reset...
18:14 Dr_Drache and redeploy.
18:14 Dr_Drache LOL
18:15 Dr_Drache MiroslavAnashkin, I do have a GUI bug though.
18:16 Dr_Drache if i take 2 nodes, and adjust the patitions, then apply.. only one applys and an error is produced.
18:18 MiroslavAnashkin Oh. please press Shift and click Reload button.
18:19 Dr_Drache I can go back in and redo it.
18:19 MiroslavAnashkin Browser may have something cached.
18:19 Dr_Drache and it's reporducable.
18:19 Dr_Drache but.. just giving you the paste.
18:19 Dr_Drache it's not a stopper at all
18:19 Dr_Drache http://paste.openstack.org/show/75003/
18:21 MiroslavAnashkin Great. It looks exactly like new bug.
18:21 Dr_Drache both nodes are 100% identical
18:21 Dr_Drache process :
18:22 Dr_Drache configure disk, choose first disk, change virtual volume to 5,120, click "use all space" for ceph, click apply
18:28 jobewan joined #fuel
19:15 Dr_Drache MiroslavAnashkin, same error
19:15 Dr_Drache it errors out after all nodes get to 100%
19:16 xarses Dr_Drache: what's the error?
19:18 Dr_Drache xarses,
19:18 Dr_Drache http://imgur.com/t4QZ7dA
19:20 Dr_Drache xarses, it does seem that the cluster is live though
19:21 xarses last 300 lines of /var/log/nailgun/app.log and /var/log/naily/naily.log
19:21 xarses please
19:21 Dr_Drache k
19:23 Dr_Drache app.log = http://paste.openstack.org/show/75015/
19:25 Dr_Drache naily.log = http://paste.openstack.org/show/75016/
19:26 Dr_Drache it also shows 2 copies of TestVM Images in a constant state of "saving"
19:31 e0ne joined #fuel
19:34 xarses ceph?
19:34 xarses how is 'ceph -s'
19:35 Dr_Drache trying to get in.
19:35 xarses and 'cinder create 1' then 'sleep3; cinder show <uuid>'
19:37 Dr_Drache fuel can't even see the nodes.
19:37 xarses they are "offline"?
19:38 Dr_Drache no routes.
19:39 xarses eh, isn't their admin IP on the fuel-network, they should have local routes
19:39 xarses 'mco ping'
19:40 Dr_Drache they come up there.
19:40 Dr_Drache weird
19:40 xarses but ssh node-x doesn't?
19:40 Dr_Drache right
19:41 xarses you can see if 'fuel node' show's the ip addresses, i cant remember if thats in 4.1 or master
19:42 Dr_Drache [root@fuel naily]# fuel node
19:42 Dr_Drache id | status | name             | cluster | mac               | roles                        | pending_roles | online
19:42 Dr_Drache ---|--------|------------------|---------|-------------------|------------------------------|---------------|-------
19:42 Dr_Drache 12 | ready  | Untitled (0B:A0) | 5       | 1e:ac:58:75:fe:4c | [u'compute', u'ceph-osd']    | []            | True
19:42 Dr_Drache 11 | ready  | Untitled (0C:DB) | 5       | 7a:b3:e5:c2:75:42 | [u'compute', u'ceph-osd']    | []            | True
19:42 Dr_Drache 10 | ready  | Untitled (ce:7d) | 5       | 3e:c6:21:0b:49:4f | [u'controller', u'ceph-osd'] | []            | True
19:44 xarses if not we can add it
19:46 xarses http://paste.openstack.org/show/75020/
19:47 Dr_Drache ok
19:51 Dr_Drache http://paste.openstack.org/show/75023/
19:52 Dr_Drache and I did notice... the openstack part of the deployment took much longer than on patch v2
19:52 Dr_Drache xarses,
19:53 xarses zero osd's
19:53 Dr_Drache yea
19:54 xarses so thats why the image upload failed
19:54 Dr_Drache should have 8, according to my last deployment
19:54 xarses thats odd. /root/ceph.log from one of them
19:55 Dr_Drache http://paste.openstack.org/show/75024/
19:56 Dr_Drache and
19:56 xarses oddness, it didn't try to provision the osd devices
19:57 Dr_Drache http://paste.openstack.org/show/75025/
19:58 Dr_Drache fuel was deploying perfectly with patch ver2, I reset my nodes, installed ver4, manually edited my pmanager.py for "nofb", cobbler sync, rebooted, and, i'm on my 4th redeploy with these errors.
20:00 xarses 'facter -p osd_devices_list' on one of the osd hosts
20:01 Dr_Drache nothing
20:01 Dr_Drache on 2 nodes
20:05 xarses are the ceph devices mounted already?
20:05 Dr_Drache I'd assume they would be.
20:05 Dr_Drache but I did nothing manually
20:06 Dr_Drache what I currently have, is what fuel gave us
20:06 xarses they shouldn't be in this case
20:08 Dr_Drache k
20:14 IlyaE joined #fuel
20:15 Dr_Drache xarses, I'd go back to ver2 if I could, LOL, just wanted the logout template back.
20:15 xarses logout template?
20:16 Dr_Drache https://bugs.launchpad.net/fuel/+bug/1284241
20:17 Dr_Drache but I cannot go back to 2 anyway, lol, no download.
20:17 xarses you didn't keep a copy of it?
20:18 Dr_Drache I downloaded it directly to the fuel master. other patches were still availble, I didn't forsee them getting purged.
20:18 Dr_Drache my mistake there for sure.
20:19 xarses fuel_partition_manager_patch_40_to_41_ver2?
20:19 xarses or 41_to_411
20:19 Dr_Drache 41 to 411
20:22 xarses hmm, I'd guess it was https://bugs.launchpad.net/fuel/+bug/1297792 that caused this oddness. I'm not sure why it changes your provision behavior this much
20:23 Dr_Drache I don't honestly know either.
20:23 Dr_Drache I went through methotically before I came here this time.
20:55 Dr_Drache xarses, I'm a few min away from needed to wrap up my day, just FYI
20:56 xarses Dr_Drache: sorry, I don't have anything useful for you I was lightly looking for https://bugs.launchpad.net/fuel/+bug/1297792 to see what it's doing but it isn't linked properly
20:57 Dr_Drache xarses, no that's fine. just didn't want to leave without saying something.
20:57 Dr_Drache I might go back to default 4.1 and manually patch the 2 things required to deploy.
22:49 mattymo joined #fuel
23:15 e0ne_ joined #fuel

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