Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-05-31

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

All times shown according to UTC.

Time Nick Message
00:22 raunak joined #fuel
00:49 Julien-zte joined #fuel
01:54 gardlt joined #fuel
02:31 gardlt joined #fuel
02:49 dilyin joined #fuel
03:13 Julien-zte joined #fuel
03:54 neophy joined #fuel
05:31 kszukielojc joined #fuel
05:40 Julien-zte joined #fuel
06:09 Zer0Byte___ joined #fuel
06:18 javeriak joined #fuel
06:21 Miouge joined #fuel
06:50 parlos_ joined #fuel
06:53 neophy joined #fuel
06:57 krypto joined #fuel
07:03 krypto i am facing strange issue with fuel6.1 and contrail plugin.Newly added compute nodes are configured with openvsiwtch and there are no contrail package installed.But still fuel webui shows status as ready.I tired to execute /etc/fuel/plugns/contrail/puppet/.....packages.pp but that exits with out any errors.Any idea about this issue
07:13 javeriak joined #fuel
07:15 kutija joined #fuel
07:39 natarej_ joined #fuel
07:44 francois joined #fuel
07:47 francois hey guys, I'm looking for information on AD/LDAP auth support in Fuel 9
07:47 francois is it gonna be based on https://github.com/openstack/fuel-plugin-ldap or something new ?
07:56 noshankus joined #fuel
07:56 parlos joined #fuel
08:02 vstoiko joined #fuel
08:07 myatsenko joined #fuel
08:10 natarej__ joined #fuel
08:11 myatsenko francois: hello,
08:11 myatsenko yes, version of ldap_plugin for Fuel9 will be based on https://github.com/openstack/fuel-plugin-ldap,
08:11 myatsenko but the code will be updated to support additional features.
08:13 omolchanov joined #fuel
08:13 francois hi myatsenko: ok thanks
08:13 francois we're planning a new deployment based on fuel 8 with AD auth
08:14 francois and I had to manually build the fuel-plugin-ldap as it doesn't seem to be officially available
08:14 francois so that would mean that upgrading from fuel 8 to 9 with this ldap plugin probably won't be too much an issue ?
08:22 elopez joined #fuel
08:42 yassine joined #fuel
08:43 HeOS joined #fuel
08:43 yassine joined #fuel
09:49 gcossu joined #fuel
09:49 gcossu left #fuel
10:03 Julien-zte joined #fuel
10:05 javeriak joined #fuel
10:23 javeriak joined #fuel
10:37 farreld joined #fuel
10:41 javeriak_ joined #fuel
10:48 permalac joined #fuel
10:53 Julien-zte joined #fuel
11:14 javeriak joined #fuel
11:16 javeriak joined #fuel
11:20 javeriak joined #fuel
11:50 Miouge_ joined #fuel
12:42 javeriak_ joined #fuel
12:43 javeriak joined #fuel
12:46 parlos joined #fuel
12:46 parlos_ joined #fuel
13:17 ezpz joined #fuel
13:58 MikeG451 joined #fuel
14:13 bklynfare joined #fuel
14:15 Julien-zte joined #fuel
14:16 bklynfare setting up a new openstack env with Fuel 9.0 iso, network verification keeps failing due to 'nodes not available via mcollective', I can see the connections on the fuel master to port 61613 but mco ping is failing, this is a fresh iso install, anyone else see this?
14:26 yassou joined #fuel
14:29 mwhahaha bklynfare: did you change the network interface from eth0 on the master?
14:29 mwhahaha you may need to adjust the iptables rules
14:31 penguinolog joined #fuel
14:31 parlos joined #fuel
14:32 bklynfare no, I didnt change the interface mwahahaha
14:37 mwhahaha Are there any errors in the mcollective logs for those nodes?
14:38 bklynfare I see this on the master: STOMP error frame sent: Message: "Bad CONNECT" Detail: "Access refused for user 'mcollective'\n"
14:38 bklynfare I tried mco ping
14:38 fedexo joined #fuel
14:38 bklynfare let me check a slave
14:39 bklynfare no errors, just info, warn debug
14:41 mwhahaha hmm is mcollective up on the master?
14:42 bklynfare yes, and I see established connections on port 61613
14:56 xarses joined #fuel
14:57 mwhahaha you could try rebooting the slaves
14:58 bklynfare ya, I'll try that, I did restart mcollective on them
15:00 mwhahaha not sure what your environment looks like, but if you say redeployed the master with slave nodes still up then the creds get changed and you need to reboot the servers to get the new ones
15:05 bklynfare joined #fuel
15:07 bklynfare joined #fuel
15:07 bklynfare ugh, I think I did redeploy the master w/o rebooting slaves, thx mwhahaha
15:07 mwhahaha np
15:32 javeriak joined #fuel
15:42 slefeuvre joined #fuel
15:42 slefeuvre Hello all,
15:43 slefeuvre I am encountering an issue with FUEL 7.0 and health check.
15:43 slefeuvre I deploy succesfully with ceph in HA mode
15:44 slefeuvre but I get 4 check failed when I launch health check (MYSQL,GALERA,RABBITMQ)
15:45 slefeuvre everythink look good in the environment so I don't understand ?
15:45 mwhahaha how many controllers?
15:45 xarses do you have 3 or more controllers?
15:45 slefeuvre 2 controller
15:45 mwhahaha you need at least 3
15:45 mwhahaha or it's not HA
15:46 slefeuvre Everything is working, I can write and replicate data without any problem ?
15:46 yassine joined #fuel
15:46 slefeuvre I don't understand why it can't work with 2 controller ?
15:46 xarses it will work with two controllers, but those tests require an odd number more than 1
15:47 xarses you are at a high risk for split brain with only 2 controllers
15:47 yassine joined #fuel
15:47 slefeuvre Yes, I know it, but 3 is better to do in a automatic way
15:48 slefeuvre I should do with 2 by doing migration manually in order to evicted split brain
15:49 xarses you can, but you have a high risk of data loss
15:49 xarses we don't configure the cluster to stop the partation with just two nodes
15:49 xarses both will keep on working
15:49 slefeuvre Yes, but the 3rd controller should be added in a future
15:50 xarses also, ceph is particular about having a quorum, and may stop everything if it doesn't have one
15:50 kutija joined #fuel
15:50 slefeuvre But the mysql is currently configure in haproxy in active / backup mode
15:51 xarses yes, as I said, the health test will fail simply because you don't have a quorum of nodes
15:51 slefeuvre But is no way to get success check when using 2 controller ?
15:51 xarses no
15:51 xarses because it's not safe
15:51 slefeuvre ok, that reply to my question
15:52 slefeuvre many thanks to your reply it will help me better ;)
15:52 slefeuvre In the same way, If i want to add a new controller, every nodes will be reinstalled ?
15:52 xarses no, it will adjust the configurations as needed
15:53 xarses every node will be touched, but not re-installed
15:53 slefeuvre ok because when I added a 2nd controller in my environement, it reinstall all nodes at the "openstack installation" process
15:54 slefeuvre yes, but so it install all openstack environment to overwrite the existing one ?
15:54 xarses yes, we have to change config on all of the nodes to add another controller
15:54 slefeuvre ok
15:55 xarses but its not reinstalling all of openstack, it re-runs puppet which then determines what, if anything needs to be re-run to make the needed changes
15:56 Miouge joined #fuel
15:58 gardlt joined #fuel
16:02 slefeuvre xarses: ok, many thanks to your help ;)
16:03 Zer0Byte__ joined #fuel
16:08 raunak joined #fuel
16:12 bklynfare is it possible to prefer an ntp server in the fuel web interface?
16:18 thumpba joined #fuel
16:19 xarses for the cluster? its on the settings page
16:22 raunak joined #fuel
16:24 thumpba joined #fuel
16:24 bklynfare ya, xarses, it looks like I can only specify ntp servers though, not the iburst or prefer options
16:25 bklynfare I can change the puppet module
16:27 xarses correct, the ui doesn't set any of the other options
16:40 bklynfare joined #fuel
16:42 bklynfare joined #fuel
16:54 jerryz joined #fuel
16:58 fritchie joined #fuel
17:01 ezpz joined #fuel
17:19 elopez joined #fuel
17:40 Miouge joined #fuel
17:45 raunak joined #fuel
17:55 javeriak joined #fuel
17:59 javeriak_ joined #fuel
18:02 ezpz joined #fuel
18:04 raunak joined #fuel
18:05 Miouge joined #fuel
18:12 ezpz_ joined #fuel
18:12 ezpz_ left #fuel
18:13 ezpz joined #fuel
18:29 Miouge joined #fuel
18:54 gardlt joined #fuel
19:06 javeriak joined #fuel
19:33 gardlt joined #fuel
19:45 fritchie joined #fuel
19:50 raunak joined #fuel
19:50 hyperscsi joined #fuel
20:09 fritchie joined #fuel
20:12 buddy joined #fuel
20:13 Guest36448 Deployed openstack with fuel8 unable to access public vip from outside the network, but able to access dashboard and sping the instances on same network.
20:14 Guest36448 instances floating ip also not accessable from outside the network
20:20 xarses seems like your network is not configured to forward that traffic
20:26 Guest36448 I have other openstack environments working fine on same networks
20:26 Guest36448 is there any thing I need to do with the iptables?
20:27 Guest36448 First time when I deployed am able to access VIP outside the network, now I redeployed again with same network am unable to access VIP
20:28 Guest36448 deployment is not consistent with same network settings
20:37 xarses no there shouldn't be any need to use iptables
20:37 xarses you can manually check that you can reach the networks gateway from each of the controllers
20:38 Guest36448 it is just 1 controller and 1 compute node
20:38 xarses and that the vrouter and haproxy namespaces (there will be one each somewhere between all of the controllers) can reach the gateway
20:38 Guest36448 am able to ping gw from both the nodes cotroller and compute
20:38 xarses and can you ping past the gateway to the next hop?
20:39 Guest36448 yes
20:39 xarses and the reverse doesn't work?
20:40 Guest36448 there is not issues accessing from MOS environment
20:40 Guest36448 no*
20:40 xarses can you ping the environments gateway from an external network?
20:40 Guest36448 yes
20:41 xarses and not the vip?
20:41 Guest36448 yes
20:41 xarses what is the subnetmask and last to octets of the vip?
20:41 Guest36448 able to ping br-ex from outside
20:42 Guest36448 192.127 255.255.255.0
20:42 xarses what's br-ex's last two octets?
20:45 xarses I'm guessing your upstream network thinks that that network is a /25, because that would be the broadcast address for that network, if br-ex is 126 or below, that explains alot.
20:45 xarses Otherwise, that address already has a mac on the subnet, and it's receiving your packets
20:45 xarses `arp -an` might help, should show incomplete
20:51 Guest36448 192.128
21:02 xarses ya, then the ip is probably taken
21:03 xarses you can try pinging out from the namespace
21:03 xarses `ip netns exec haproxy ping`
21:03 xarses but it will probably fail
21:04 HeOS joined #fuel
21:08 Guest36448 yes it failed connect: Network is unreachable
21:09 Guest36448 and I dont see pubicl vip in haproxy ns
21:09 Guest36448 ip netns exec haproxy ip a
21:09 Guest36448 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
21:09 Guest36448 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
21:09 Guest36448 inet 127.0.0.1/8 scope host lo
21:09 Guest36448 valid_lft forever preferred_lft forever
21:09 Guest36448 inet6 ::1/128 scope host
21:09 Guest36448 valid_lft forever preferred_lft forever
21:09 Guest36448 25: b_management: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
21:09 Guest36448 link/ether 12:97:1a:07:96:8b brd ff:ff:ff:ff:ff:ff
21:09 Guest36448 inet 172.24.90.41/24 scope global b_management
21:09 Guest36448 valid_lft forever preferred_lft forever
21:09 Guest36448 inet6 fe80::1097:1aff:fe07:968b/64 scope link
21:09 Guest36448 valid_lft forever preferred_lft forever
21:09 Guest36448 27: hapr-ns: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
21:09 Guest36448 link/ether 6e:1e:ae:b4:48:02 brd ff:ff:ff:ff:ff:ff
21:09 Guest36448 inet 240.0.0.2/30 scope global hapr-ns
21:09 Guest36448 valid_lft forever preferred_lft forever
21:09 Guest36448 inet6 fe80::6c1e:aeff:feb4:4802/64 scope link
21:09 Guest36448 valid_lft forever preferred_lft forever
21:11 xarses hrm, what does pcs tell us about the public_vip resource?
21:12 Guest36448 vip__public    (ocf::fuel:ns_IPaddr2): Started node-10.domain.tld
21:12 Guest36448 Clone Set: clone_ping_vip__public [ping_vip__public]
21:12 Guest36448 Started: [ node-10.domain.tld ]
21:13 xarses huh, thats weird
21:15 Guest36448 now am able to ping it
21:15 Guest36448 ip netns exec haproxy ping 130.3.192.127
21:15 Guest36448 PING 130.3.192.127 (130.3.192.127) 56(84) bytes of data.
21:15 Guest36448 64 bytes from 130.3.192.127: icmp_seq=1 ttl=64 time=0.053 ms
21:15 Guest36448 64 bytes from 130.3.192.127: icmp_seq=2 ttl=64 time=0.038 ms
21:31 mattgriffin joined #fuel
22:10 gardlt joined #fuel
22:10 Julien-zte joined #fuel
22:44 aslaen joined #fuel
22:44 aslaen joined #fuel
22:50 buddy joined #fuel
22:56 sbfox joined #fuel

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