Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-03-21

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

All times shown according to UTC.

Time Nick Message
00:42 skamithi13 joined #fuel
00:50 vsedelnik joined #fuel
01:30 Liuqing joined #fuel
01:51 vsedelnik joined #fuel
02:26 fedexo joined #fuel
02:50 gongysh joined #fuel
02:52 vsedelnik joined #fuel
03:40 xarses_ joined #fuel
03:42 l4ng1t joined #fuel
03:53 vsedelnik joined #fuel
04:24 treaki_ joined #fuel
04:53 vsedelnik joined #fuel
04:54 Jeffrey4l__ joined #fuel
05:02 vsedelnik joined #fuel
05:34 skamithi13 joined #fuel
05:38 skamithi13 joined #fuel
05:44 fzhadaev joined #fuel
05:51 Liuqing joined #fuel
06:10 vsedelnik joined #fuel
06:13 gongysh joined #fuel
06:36 Jeffrey4l__ joined #fuel
07:10 gongysh joined #fuel
07:26 vsedelnik joined #fuel
07:30 magicboiz joined #fuel
07:47 HeOS joined #fuel
07:56 pbelamge joined #fuel
08:00 akaszuba joined #fuel
08:09 pbelamge joined #fuel
08:20 obcecado joined #fuel
08:40 permalac joined #fuel
08:55 Jeffrey4l__ joined #fuel
09:05 permalac joined #fuel
09:10 Alcest joined #fuel
09:12 neilus joined #fuel
09:23 izinovik_ joined #fuel
09:24 SergK joined #fuel
09:24 dkaigarodsev joined #fuel
09:25 pasquier-s joined #fuel
09:25 elemoine joined #fuel
09:26 fBigBro joined #fuel
09:28 gongysh joined #fuel
09:34 samuelBartel joined #fuel
09:50 neilus joined #fuel
09:50 pbelamge joined #fuel
09:52 e0ne joined #fuel
11:07 neilus joined #fuel
11:15 Liuqing joined #fuel
11:51 vvalyavskiy joined #fuel
11:51 vvalyavskiy left #fuel
11:52 vvalyavskiy joined #fuel
12:03 Liuqing joined #fuel
12:06 neilus joined #fuel
12:07 skamithi14 joined #fuel
12:15 neilus joined #fuel
12:24 Liuqing joined #fuel
12:43 acca joined #fuel
13:14 Alcest joined #fuel
13:22 spyzalski joined #fuel
13:34 permalac joined #fuel
13:38 BobBall joined #fuel
13:39 BobBall How do most people deploy their Fuel master so it can be accessed by an external network?  https://bugs.launchpad.net/fuel/+bug/1550221 seems to stop us using a second interface to give external access to the fuel master
13:42 acca left #fuel
13:45 acca joined #fuel
13:45 acca left #fuel
13:51 rmoe joined #fuel
14:14 kozhukalov BobBall: seems there has been added yet  another check when setting eth1 dhcp to true
14:16 kozhukalov rmoe: could you please help with this?
14:17 kozhukalov as far as I remember you made some changes related to second master node interface
14:20 kozhukalov BobBall: as a fast workaround you could edit /etc/fuel/astute.yaml file manually to enable dhcp on eth1
14:22 BobBall Just astute.yaml?  I assumed we'd need to modify the local network-scripts for eth1 as well to set it up? Or do they dynamically read from astute.yaml?
14:22 kozhukalov give a second to check it out
14:24 kozhukalov btw, could you please check the log file /var/log/fuelmenu.log
14:25 kozhukalov if there is a line like 'No DHCP servers found. Cannot enable DHCP'
14:25 kozhukalov ?
14:26 kozhukalov or something like this "dhcp_checker failed to check on eth1"
14:27 kozhukalov as far as I can see there is a check if dhcp server is available and working https://github.com/openstack/fuel-menu/blob/master/fuelmenu/modules/interfaces.py#L159-L173
14:28 zefciu joined #fuel
14:29 BobBall dhcp works from the console, just not fuelmenu
14:29 BobBall so if I run dhclient manually everything seems to be setup right but fuelmenu just stalls waiting on puppet to return (according to the log when I ctrl+c)
14:37 kozhukalov yes, according to what I see, fuelmenu tries to configure interfaces using puppet module l23network
14:38 xarses_ joined #fuel
14:38 kozhukalov and when you try to run dhclient manually, how long does it take to get an IP?
14:39 kozhukalov there is 5 secs  timeout
14:40 Jeffrey4l__ joined #fuel
14:41 kozhukalov as a fast work around you could configure eth1 manually in fuelmenu using the same configuration that you get when run dhclient manually
14:43 kozhukalov and once the master node is deployed you could reconfigure eth1 manually so it uses dhcp
14:52 skamithi13 joined #fuel
14:52 vsedelnik joined #fuel
14:56 omolchanov joined #fuel
15:23 vvalyavskiy left #fuel
15:29 l4ng1t joined #fuel
15:30 claflico joined #fuel
15:31 kur1j_ joined #fuel
15:32 Kdecherf is there a way to define bonding interfaces in network template yaml?
15:35 kur1j_ Deploying a controller on as a kvm. Physical host is centos7. Fuel deploys the Ubuntu correctly, however it fails when trying to fetch archives as it cannot initiate connection to archive.ubuntu.com.
15:38 kur1j_ however I can login to the Ubuntu box from the fuel master node and ping Google initially. But about 7 minutes into the open stack deployment (I'm guessing after firewalls and stuff are configured) I can no longer ping google.com
15:38 kur1j_ this same exact network configuration works on a physical node
15:39 skamithi13 joined #fuel
15:43 mwhahaha kur1j_: does the network verification work?
15:44 kur1j_ can't deploy
15:44 kur1j_ so that can't work.
15:44 mwhahaha you can do network verification prior to the deploy
15:45 mwhahaha and if that fails, your network configuration is probably wrong
15:45 mwhahaha that's probably why it's failing because it reconfigured with your provided network settings and can no longer reach the internet
15:45 mwhahaha prior to deployment, all the traffic is routed through the master
15:45 kur1j_ I can use a physical machine with these same settings and it deploys successfully
15:46 kur1j_ it's just within the kvm environment it seems
15:46 mwhahaha i would assume it's a configuration error with your kvm setup
15:47 mwhahaha all my dev environments are kvm instances and i don't have such issues
15:47 kur1j_ is there anything special I need to do on the host os?
15:47 mwhahaha shouldn't be
15:47 kur1j_ I turned off selinux and the firewalld
15:48 mwhahaha check the arp table on you switch?
15:48 kur1j_ are you using bridging?
15:48 mwhahaha yea
15:49 kur1j_ centos7 for host?
15:49 mwhahaha no i'm using debian
15:49 kur1j_ Hmm
15:49 kur1j_ do you think having fuel within a kvm would cause issues?
15:49 mwhahaha nope
15:50 mwhahaha all our testing is done with such a configuration
15:50 mwhahaha or at least a lot of it
15:52 kur1j_ ok I'm going to try this again
15:53 kur1j_ for like the 300th time haha
15:53 gariveradlt joined #fuel
15:55 slo_ joined #fuel
15:57 kur1j_ mwhahaha: network verification configured correctly
15:57 kur1j_ just ran it between two modes
15:57 kur1j_ nodes*
15:57 mwhahaha ok so it's probably something with the kvm configuration
15:58 kur1j_ the host or the actual kvm?
15:58 kur1j_ I don't get what other settings I can pick
15:59 kur1j_ I just add 3 network adapters to the system
15:59 kur1j_ and that's it
15:59 mwhahaha the kvm host
15:59 mwhahaha for example, https://bugs.centos.org/view.php?id=5526 see the last comment
16:00 v1k0d3n joined #fuel
16:01 kur1j_ Hmm
16:02 kur1j_ that could def be it
16:02 kur1j_ I'm going to try it with e1000 device
16:03 mwhahaha yea my env uses e1000
16:05 kur1j_ I'm going to be so damn happy if this works
16:05 kur1j_ been dealing with it for like 2 days
16:12 slo_ mwhahaha: can you maybe tell me if is possible to upgrade to MOS 7, if I have openstack installed with vanila Fuel 6: https://wiki.openstack.org/wiki/Fuel  ?
16:12 slo_ or are MOS updates only usable with MOS 6.1 distribution?
16:13 mwhahaha i want to say yes, but ogelbukh_ is more familiar with upgrades
16:16 slo_ aha ok
16:16 claflico joined #fuel
16:17 slo_ ogelbukh_:if you possible know the answer to this question, i would be glad
16:18 mwhahaha do you just want to upgrade the fuel software or the environment?
16:18 slo_ since mirantis , as said, has some hardened packages, that are possible not part of the opensource or not there yet, i am skeptical regarding updates...
16:18 slo_ fuel and the environemnt
16:18 slo_ with newer version of openstack
16:19 mwhahaha upgrading the environments was experimental for 7.0 and onlyf rom 6.1 to 7, https://docs.mirantis.com/openstack/fuel/fuel-7.0/operations.html#upgrade-environment-experimental
16:19 mwhahaha i'm pretty sure you can upgrade the fuel software though
16:20 slo_ aha so version of MOS 8 will not have an upgrade from MOS 7 package?
16:20 mwhahaha not sure about 8
16:21 kur1j_ mwhahaha: :( same issue. as soon as fuel configured the vip public stuff it stopped working
16:21 mwhahaha kur1j_: is your public gateway pingable?
16:21 kur1j_ not any more
16:21 mwhahaha was it?
16:21 kur1j_ I was on the system watching logs
16:21 kur1j_ yes
16:22 kur1j_ I was pinging entire time
16:22 mwhahaha well you'll need to troubleshoot the bridges
16:22 kur1j_ as soon as it configured vip public it stopped
16:22 Kdecherf is there a way to define bonding interfaces in network template?
16:23 ggolin hello all
16:24 ggolin so is the intent of having puppet installed in the env to also be able to manage some aspects of the system via puppet post installation? for example id like to add a user at a later time. or do i have to plan it all in advance?
16:25 mwhahaha Kdecherf: there should be a way
16:25 mwhahaha ggolin: no we don't spin up a puppet master or anything, we're working on such features for 9/10
16:25 mwhahaha not necessarily puppetmaster, but a way to manage the cloud after installation
16:27 mwhahaha Kdecherf: here was a bug about it, but it has an example template in the bug description https://bugs.launchpad.net/fuel/+bug/1476188
16:27 Kdecherf mwhahaha: yep i just saw it, thx
16:27 claflico joined #fuel
16:32 ggolin mwhahaha: so you're saying post install im on my own as far as managing the servers goes
16:33 skamithi13 joined #fuel
16:33 mwhahaha as far as additional configuration not available via fuel/plugins and other software, yes
16:37 kur1j_ mwhahaha: ok so I deleted v_vrouter_pub and v_public from my br-ex bridge and now I can ping my gateway again
16:38 BillC joined #fuel
16:39 mwhahaha hmm ok well you'll need those so i guess you'll have to figure out why those break your network connectivity
16:39 kur1j_ ok strange...
16:39 kur1j_ they got added back automatically
16:40 kur1j_ but I can still ping out to the gateway
16:40 magicboiz joined #fuel
16:41 kur1j_ before I deleted them manually I could not ping out
16:41 kur1j_ this is all after the node errored out in the fuel ui
16:43 ggolin btw on cisco usc you have to configure mtu 9216 at the CoS level for mtu 9000 to work. configuring mtu 9000 at the vif level is not enough.
16:44 kur1j_ ggolin: me?
16:45 ggolin no, not you, kur1j_ :)
16:46 ggolin jsut a general remark, i know people search the irc logs. this detail took me some time to figure out.
16:49 kur1j_ fair enough
16:50 kur1j_ I've found some stuff there so I know :)
16:55 claflico joined #fuel
17:07 kur1j_ ok I don't like this at all...I changed nothing...absolutely nothing...reset the environment and deployed again...as soon as the machine started installing openstack I started pinging the gateway and let it run the entire time...it looks like it has passed the part it was failing on earlier
17:09 kur1j_ I was just wanting to watch exactly where it was failing running ping and watching debug log...but so far hasn't failed yet and it's still pinging
17:09 kur1j_ what in the hell
17:12 ggolin are you still testing with a bunch of switches plugged into all these interfaces all over the place?
17:12 ggolin kur1j_:
17:14 kur1j_ ggolin: nope :). I went and setup the vlan trucking
17:14 kur1j_ trunking
17:15 kur1j_ i was able to succesfully deploy this to a physical node yesterday...finally
17:15 ggolin im just wondering if its less crazy than what you had going on friday (got your messages this morning with a link to your network diagram)
17:15 kur1j_ then today was trying to deploy it to a KVM and was having issues with the network dropping off when it setup the router and vip public bridges
17:16 kur1j_ i setup the logging to watch it on fuel puppet and start pinging the gateway to get a better idea of where in the deployment it fails
17:17 kur1j_ but it never failed...
17:17 kur1j_ it's at like 26% complete on the openstack deploy
17:19 pbrzozowski joined #fuel
17:19 ggolin watch the puppet logs, it might be retrying something that keeps failing
17:19 kur1j_ that's what I was doing
17:19 kur1j_ but it finished it
17:19 kur1j_ it was losing connection, failing on connecting to get to Ubuntu site for updates
17:20 kur1j_ it's way past that now bc it's setting up all the services
17:20 kur1j_ and I can still ping
17:21 kur1j_ problems like this annoy the ever living piss out of me
17:21 kur1j_ lol
17:21 kur1j_ debug th problem? nope no problem
17:21 kur1j_ not watching for it? error
17:49 rmoe joined #fuel
18:01 kur1j_ yup, deployed correctly...wtf
18:01 kur1j_ now for the real test...can I get to horizon
18:03 kur1j_ nope
18:07 ggolin times out?
18:07 ggolin did you deploy an ha setup?
18:07 kur1j_ on the controller node I can get to the horizon ip
18:08 kur1j_ ggolin: everything is default
18:08 ggolin one controller then
18:08 kur1j_ yup
18:08 skamithi14 joined #fuel
18:08 kur1j_ and when it finished deploying it got 192.168.1.3
18:08 ggolin does that install haproxy btw?
18:08 kur1j_ yes looks like it
18:09 ggolin go look in its config which ip it binds to
18:09 kur1j_ running ip netns exec haproxy bash
18:10 neilus joined #fuel
18:10 kur1j_ looking at that namespace I see management and public
18:10 kur1j_ public is 192.168.1.3
18:11 kur1j_ b_public*
18:11 kur1j_ from within that namespace I cannot ping the gateway
18:12 kur1j_ from the host (regular namespace) I can ping the 192.168.1.1 gateway
18:12 kur1j_ routes look correct
18:15 kur1j_ in /etc/haproxy/015-horizon.cfg I see "bind 192.168.1.3"
18:17 kur1j_ default route is "default via 192.168.1.1 dev b_public metric 10"
18:33 magicboiz joined #fuel
18:40 kur1j_ .
18:49 skamithi13 joined #fuel
18:50 skamithi13 joined #fuel
18:55 e0ne joined #fuel
18:57 magicboiz joined #fuel
19:04 gariveradlt joined #fuel
19:07 magicboiz joined #fuel
19:20 kur1j_ ok yeah something is going screwy with v_vrouter_pub bridge
19:25 poseidon1157 joined #fuel
19:25 poseidon1157 Hi all.  I'm having a deployment issue
19:25 poseidon1157 nodes can reach the internet correctly, but I get a hash mismatch error
19:26 poseidon1157 E: Failed to fetch http://<fuel>:8080/liberty-8.0/ubuntu/x86_64/pool/main/p/python-netaddr/python-netaddr_0.7.17-1~u14.04+mos1_all.deb Hash Sum mismatch
19:26 poseidon1157 Is this a known bug?
19:26 kur1j_ do a network check?
19:26 poseidon1157 Net check works
19:27 poseidon1157 Environments: Verification succeeded. Your network is configured correctly.
19:27 kur1j_ can you ping outside world from box?
19:27 poseidon1157 yeah
19:27 kur1j_ node your are deploying too?
19:28 poseidon1157 from fuel master as well as the node
19:28 kur1j_ hmm change the values where you are getting sources?
19:28 poseidon1157 I have not
19:29 kur1j_ sounds like a validation failed
19:29 rmoe joined #fuel
19:29 kur1j_ not sure though.
19:30 poseidon1157 http://snag.gy/0Tc7f.jpg
19:30 poseidon1157 Standard MOS 8.0 installation
19:30 poseidon1157 Network connectivity checks out
19:31 poseidon1157 I work for mirantis btw.  In the training dept.
19:31 poseidon1157 This is my home lab
19:31 angdraug joined #fuel
19:31 kur1j_ I'm just a random person lol
19:31 ggolin can you wget that file?
19:31 ggolin poseidon1157:
19:31 poseidon1157 wget which file?
19:32 ggolin the one that is mismatching on hash sum
19:32 poseidon1157 k, let me try
19:32 ggolin from the target node, yes?
19:32 ggolin poseidon1157:
19:32 poseidon1157 yes
19:32 poseidon1157 I get binary back from curl
19:33 ggolin you get the debian  package?
19:33 poseidon1157 0acfd46d0dc8d6d127eb11fdf6dc2970  python-netaddr_0.7.17-1~u14.04+mos1_all.deb
19:33 poseidon1157 Yes
19:34 gariveradlt joined #fuel
19:34 ggolin this could be caused by a blip in the network
19:34 ggolin does this happen consistently?
19:34 ggolin or for random packages?
19:34 poseidon1157 I'm confident that this is either a) an issue with the mirantis repo, or more concerningly b) an issue with the hash value on the 8.0 image
19:34 poseidon1157 Just that one
19:34 poseidon1157 And this is my second install
19:35 ggolin then i agree with your assessment
19:35 poseidon1157 deploys that worked under 7.0
19:35 poseidon1157 Is there a hack I can do in the interrim?
19:36 poseidon1157 Anyone know where those are stored?  Is it just dpkg?
19:39 ggolin run a find for that filename on the master
19:39 ggolin oh i bet its in one of the containers?
19:39 poseidon1157 Yeah, I just tried under /etc/puppet
19:39 ggolin i dont think you will find it there. see what binds to port 8080 on the master and go from there.
19:39 ggolin lsof -i:8080
19:40 ggolin poseidon1157: wait a minute did you just edit the url or was it showing up that way in the error message? http://<fuel>?
19:41 poseidon1157 Yeah, it's my local ip
19:41 ggolin ah ok
19:41 poseidon1157 It points to the correct server
19:41 poseidon1157 I can dump some logs to a formal ticket if it's necessary
19:42 * ggolin does not work for mirantis, just a note.
19:42 poseidon1157 thanks
19:43 poseidon1157 Oh, hey v1k0d3n
19:47 mwhahaha poseidon1157: did you setup a local mirror?
19:48 poseidon1157 I did not
19:48 poseidon1157 Paste above should show that there's none referenced
19:48 mwhahaha i've not seen that error before as it relates to fuel provided packages
19:49 mwhahaha does it happen everytime you installed a fuel master or did you only try once?
19:50 poseidon1157 mwhahaha I've tried it twice now
19:51 poseidon1157 mwhahaha In either case the install works fine, but the deploy breaks
19:51 poseidon1157 I actually ran into one other issue.  I'm using a cheap i7, and I had to turn ACPI off in the kernel boot parameters
19:51 poseidon1157 But otherwise the install worked fine
19:52 v1k0d3n poseidon1157: hey
19:52 mwhahaha i've got a different md5sum for my python-netaddr package
19:52 mwhahaha it's from a 9.0 iso, but it's 4ca686f932dde6269848e2e465f0a235  /var/www/nailgun/liberty-9.0/ubuntu/x86_64/pool/main/p/python-netaddr/python-netaddr_0.7.17-1~u14.04+mos1_all.deb
19:52 mwhahaha in 8.0 it'd be under the nailgun docker container i believe
19:52 poseidon1157 I'll check
19:52 poseidon1157 8080 was binding to nginx
19:53 poseidon1157 also http://snag.gy/ezNbP.jpg
19:53 mwhahaha yea that fronts nailgun
19:53 poseidon1157 k, thanks
19:56 poseidon1157 found the file
19:57 poseidon1157 Don't suppose I can just delete it?
20:03 mwhahaha no
20:03 mwhahaha you need it
20:03 mwhahaha so either you have a bad copy or you need to regenerate the hash info
20:04 kur1j_ would there be a reason why I couldn't get to the gateway from the haproxy or vrouter namespace?
20:05 poseidon1157 Anyone know where I can download that file?
20:05 magicboiz joined #fuel
20:05 poseidon1157 And I would need to regenerate the hash on each node, right?
20:06 poseidon1157 I'm looking at the nailgun container, and the node is receiving the file in full
20:06 poseidon1157 I don't know where to find the expected hash value though
20:09 poseidon1157 removing the file gives me a 404
20:09 poseidon1157 So I have to have it
20:09 mwhahaha no just the nailgun container
20:09 poseidon1157 I'm in the nailgun container
20:09 poseidon1157 So here's the issue: the server may have a bad file, or the bootstrap image may expect a bad hash value
20:10 poseidon1157 I don't know where to solve this on either end
20:10 poseidon1157 I can't find the expected hash on the node to be deployed.  I checked in /var/lib/dpkg/info, but to no avail
20:10 poseidon1157 And the server doesn't seem to show me where it sourced the file from
20:28 xarses kur1j: you should be able to to reach them given the public gateway IP was reachable during deployment
20:28 mwhahaha poseidon1157: you can try re downloading the deb from http://mirror.fuel-infra.org/mos-repos/ubuntu/8.0/pool/main/p/python-netaddr/
20:33 poseidon1157 tyvm
20:37 poseidon1157 fixed that part
20:37 poseidon1157 :)
20:38 kur1j_ xarses: that's the thing, when I do a deploy of a single controller it will sometimes deploy but can't get to the horizon ip. other times it will just fail to deploy at all.
20:39 kur1j_ erroring out saying it can't connect to the Internet
20:42 kur1j_ yup this time going to fail, as soon as it configured the vip_router stuff Internet stopped working on the node being deployed
20:43 ggolin could you be setting up a colliding ip address for your external network?
20:44 kur1j_ vip_vrouter_pub stuff breaks
20:44 kur1j_ I don't think so.
20:45 kur1j_ there are 3 boxes I know of on the 192.168.1.0/24. network.
20:45 mwhahaha are you using the default networks?
20:46 kur1j_ fuel, which is .6, .1 the gateway, my laptop .9
20:46 mwhahaha so vrouter is going to endup with .1
20:46 mwhahaha which would explain why it dies
20:46 * ggolin chuckles
20:47 kur1j_ wtf
20:47 kur1j_ why?
20:47 kur1j_ the gateway in my public is 192.168.1.1...which is my external network gateway
20:47 mwhahaha well vrouter_pub will be .2
20:48 kur1j_ why would that be a problem?
20:48 mwhahaha sec i'm getting there, trying to cross reference my environment
20:48 kur1j_ sure
20:48 mwhahaha need to figure out which networks go where :D
20:48 kur1j_ haha fair enough
20:48 kur1j_ yeah this stuff is confusing as hell lol
20:48 kur1j_ the network side of things
20:49 mwhahaha so in the vrouter network namespace, it creates a vrouter_pub which is on the public network (using .2)
20:49 mwhahaha it also creates a vrouter which is the .1 of the management network
20:49 mwhahaha the default gateway for vrouter network namespace would be the .1 of the gateway
20:50 kur1j_ my management network is 10.0.20.0/24
20:50 mwhahaha so from within the vrouter namespace (ip netns exec bash) can you ping/traceroute anything?
20:50 kur1j_ no
20:50 mwhahaha is this vlan or tunneling?
20:50 kur1j_ supposed to be tunneling
20:51 kur1j_ tunneling as in neutron VXLAN
20:52 mwhahaha what does your arp table look like in the namespace?
20:52 kur1j_ it's empty
20:52 kur1j_ I can't check it now bc I reset it and was trying something else
20:52 kur1j_ it failed though
20:53 kur1j_ so the gateway I put in the fuel UI is that not my external routers gateway ip?
20:53 mwhahaha yea it should be
20:54 mwhahaha are you running dhcp on that existing network?
20:54 kur1j_ I mean it kinda makes sense that there is something funny going on with collisions bc sometimes it will deploy and other times it will fail
20:54 kur1j_ no
20:55 mwhahaha did you adjust the public ips to start after your .9?
20:55 kur1j_ no it's at .2 to 100
20:55 kur1j_ good point
20:57 kur1j_ going to try that
20:57 ggolin yeah start at the end of your routers dhcp pool +1
20:57 kur1j_ ggolin: no dhcp on this subnet
20:58 kur1j_ but would be surprised that this works
20:58 kur1j_ I never saw anything assigning anything to 9
20:58 kur1j_ or conflicting
20:59 ggolin this is why i suggested you do the testing with vms on one server, confined to a virtual environment.
20:59 kur1j_ this is on a van
20:59 kur1j_ vm
20:59 kur1j_ well all of its hosting on the vm
21:01 gariveradlt joined #fuel
21:05 vsedelnik joined #fuel
21:06 kur1j_ so slightly different...failed trying to ping the gateway
21:07 kur1j_ it keeps repeating "executing ping -q -c 1 -W 3 192.168.1.1"
21:07 kur1j_ I changed nothing other than starting ip for the public space
21:08 kur1j_ reset environment and added more
21:08 kur1j_ node*
21:08 kur1j_ deployed
21:08 kur1j_ I don't get this
21:09 mwhahaha that check occurs during the network configuration
21:10 mwhahaha basically it's a ping check to make sure we can ping the gateway after configuring the network interfaces
21:10 kur1j_ gateway is correct in route -n
21:11 kur1j_ what would it be changing that causes the network is flake out like that?
21:11 kur1j_ I see nothing in the setting wrong for network related stuff
21:11 kur1j_ route looks fine
21:12 mwhahaha mtu issues? switch configuration?
21:12 mwhahaha networking is hard and annoying to trouble shoot :D
21:12 neilus joined #fuel
21:13 mwhahaha i'd recommend just starting from the network namespace and start troubleshooting each connection
21:13 kur1j_ deleted bridge p_ff798dba-0 and now I can ping gateway...
21:14 mwhahaha was that on br-ex?
21:15 kur1j_ Yes
21:15 kur1j_ the vrouter stuff wasn't even setup yet
21:16 mwhahaha so that's the patch between the br-ex and br-floating
21:17 kur1j_ how would I debug that?
21:18 kur1j_ I mean it should be correct
21:18 mwhahaha i found that by grepping it out of /var/log/puppet.log on the host
21:19 mwhahaha so on my host...http://paste.openstack.org/show/491360/
21:20 kur1j_ oddly enough not on my controller node
21:21 mwhahaha you'll end up with different hash names
21:21 mwhahaha that's just an example from my 9.0 env
21:21 kur1j_ that's the name of mine
21:22 kur1j_ it's in /etc/network/interfaces/ifcfg-p_ff789dba-0
21:23 kur1j_ I'm about to dump centos and install these vms on esxi
21:24 kur1j_ I can't tell if that's causing the issue, KVM configuration, or fuel itself
21:28 v1k0d3n joined #fuel
21:35 magicboiz joined #fuel
22:17 skamithi13 joined #fuel
22:18 skamithi13 joined #fuel
22:52 fatdragon joined #fuel
22:59 poseidon1157 left #fuel
23:08 rmoe joined #fuel
23:14 rmoe joined #fuel
23:32 rmoe joined #fuel
23:33 rmoe joined #fuel

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