Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-11-15

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

All times shown according to UTC.

Time Nick Message
00:27 Julien-zte joined #fuel
00:33 BobBall_AWOL joined #fuel
00:52 wolf joined #fuel
01:02 Julien-zte joined #fuel
01:29 gongysh2 joined #fuel
02:12 cartik joined #fuel
02:48 ilbot3 joined #fuel
02:48 Topic for #fuel is now Fuel 9.0 (Mitaka) https://www.fuel-infra.org/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
02:58 zimboboyd joined #fuel
03:28 aneliubin joined #fuel
05:11 cartik joined #fuel
05:36 Julien-zte joined #fuel
06:25 Vijayendra joined #fuel
07:09 Miouge joined #fuel
07:15 vkulanov joined #fuel
07:18 YuYangWang joined #fuel
08:11 magicboiz joined #fuel
08:11 magicboiz left #fuel
08:30 Zer0Byte joined #fuel
10:15 Julien-z_ joined #fuel
10:40 krypto joined #fuel
10:43 krypto hello all i have 3 controller nodes deployed using fuel 7 ,how many nodes failure can this setup handle,i was in an assumption that even with atleast one node my cluster will be operational but with only one controller up my vips are all in stopped state
11:24 kelepirci hi krypto. You can loose only one controller node.
11:26 kelepirci because of corosync policy.
11:26 krypto kelepirci thanks,by bringing up second controller vip became online.Can you tell me where is the limitaion of one node handling whole cluster.In a single controller set up everything is fine.
11:27 kelepirci if you lose two nodes in 3 nodes cluster, pacemaker and corosync will stop working.
11:29 krypto okk thanks
11:30 kelepirci but if your env has single controller node, you do not have to worry about HA
11:31 krypto for 2 nodes failure do i need 6 nodes in cluster
11:33 kelepirci actually in that case, you will need 5 controller node cluster.
11:33 krypto ahh right next odd number :)
11:33 kelepirci than you can loose two controller nodes. and if you need to extend your controller, it must be odd number.
11:42 krypto Thanks kelepirci is corosync+pacemaker solution better than keepalive+vrrp? i guess with keepalive+vrrp 2 node failure can be handled in a 3 nodes cluster
11:43 kelepirci actually problem is not about corosync and pacemaker. problem is galera cluster. galera cluster requite at least 3 nodes to be ha
11:46 krypto joined #fuel
11:46 krypto joined #fuel
11:59 [HeOS] joined #fuel
12:05 abalobashina joined #fuel
12:06 abalobashina Hi! Please, review https://review.openstack.org/#/c/388750/ and  https://review.openstack.org/#/c/397590/
12:06 HeOS joined #fuel
12:23 vkulanov joined #fuel
12:27 krypto kelepirci so if i decouple haproxy role to 3 different nodes then can corosync pacemaker handle 2 nodes failure in a 3 nodes cluster
12:34 kelepirci pacemaker and corosync can with some custom config. but galera cluster can not.
12:40 krypto okkk
13:19 [HeOS] joined #fuel
13:40 Tocko joined #fuel
14:16 AlexAvadanii packetary folks: please review https://review.openstack.org/#/c/392936/ + https://review.openstack.org/#/c/392937/
14:17 AlexAvadanii and https://review.openstack.org/#/c/392887/, but that is lower prio
14:32 kutija joined #fuel
14:41 Miouge Any word on Fuel 10.0? I see it as latest stable on fuel-infra.org but not on Mirantis.com?
14:59 DeMiNe0 joined #fuel
15:03 Julien-zte joined #fuel
15:06 Seafire joined #fuel
15:07 Seafire Hello everybody, is there a way to force the reset of a Fuel environment? I get the following error from the web GUI: A server error has occurred. Please contact your system administrator.
15:33 HeOS_ joined #fuel
15:36 GMAzrael joined #fuel
15:37 GMAzrael Anyone familiar with enabled a DHCP server on a network and bypassing neutron DHCP?
15:37 [HeOS] joined #fuel
15:37 GMAzrael /s/enabled/enabling
15:51 HeOS joined #fuel
15:54 krypto joined #fuel
15:59 [HeOS] joined #fuel
16:08 [HeOS] joined #fuel
16:13 chadmo joined #fuel
16:15 [HeOS] joined #fuel
16:15 aneliubin joined #fuel
16:20 Miouge GMAzrael: can’t they co-exist as long as you are using MAC reservations in both DHCPs?
16:20 GMAzrael Suprisingly, you can. DHCP can be disabled on any subnet you create, but OVS Port security needs to be disabled to prevent the DHCP spoofing.
16:21 GMAzrael Just figured it out in the last 15 minutes or so
16:21 GMAzrael The side effect is that security groups are then disabled
16:23 GMAzrael but this network is internal only, so we are not concerned
16:32 HeOS joined #fuel
16:33 DeMiNe0 joined #fuel
16:44 DeMiNe0 joined #fuel
16:52 DeMiNe0 joined #fuel
17:08 DeMiNe0 joined #fuel
17:26 Julien-zte joined #fuel
17:27 Julien-zte joined #fuel
17:28 Julien-zte joined #fuel
17:30 bookwar joined #fuel
17:41 DeMiNe0 joined #fuel
18:34 bookwar1 joined #fuel
19:05 Miouge joined #fuel
19:14 bookwar joined #fuel
19:39 Miouge joined #fuel
20:08 Miouge joined #fuel
20:53 Miouge joined #fuel
21:10 Miouge joined #fuel
21:12 AlexAvadanii joined #fuel
21:50 consultron joined #fuel
22:52 consultron joined #fuel
23:14 consultron joined #fuel
23:19 consultron joined #fuel

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