5 May 2018 |
@tiny6996:matrix.org | I thought it died | 14:24:46 |
TimePath | Maaaaybe | 23:52:21 |
7 May 2018 |
@tiny6996:matrix.org | We can keep it alive | 13:29:18 |
8 May 2018 |
TimePath | You think eventually people will notice the unreads? | 08:13:04 |
TimePath | Or do we need actual content? Lol | 08:13:22 |
@tiny6996:matrix.org | lol I can start gushing about proxmox does that count as content? | 14:01:37 |
9 May 2018 |
TimePath | the heck is proxmox? | 02:18:50 |
@tiny6996:matrix.org | it is a open source hypervisor based on debian that does both full virtual machines with kvm and machine containers with lxc | 04:48:00 |
TimePath | Sounds neat if you can't be bothered doing it yourself | 08:46:55 |
@tiny6996:matrix.org | It is and it makes things like ha and ceph very approachable | 13:24:23 |
10 May 2018 |
@tiny6996:matrix.org | how hard is it to setup lxd or kvm to HA? | 14:35:20 |
@tiny6996:matrix.org | how hard is it to setup lxc or kvm to Failover? | 14:35:35 |
gucky | @tiny6996:matrix.org: do you mean the vms themself or do you mean the application inside? | 14:39:25 |
@tiny6996:matrix.org | The vms themselves | 15:01:20 |
@tiny6996:matrix.org | Ex the physical host dies and the vms inside move somewhere else | 15:02:06 |
gucky | @tiny6996:matrix.org: you can setup proxmox. Www.proxmox.com | 15:14:01 |
@tiny6996:matrix.org | that is what I did and it is a pretty straight forward thing I was wondering what TimePath ment by sounds neat if you can't be bothered to do it yourself | 15:19:33 |
@tiny6996:matrix.org | like is it really that easy to setup HA on KVM or LXC without something like ovirt or proxmox? | 15:20:01 |
| * gucky @tiny6996:matrix.org you can testca 3 node cluster with pacemaker and http://www.linux-ha.org/wiki/VirtualDomain_(resource_agent) | 15:26:28 |
gucky | @tiny6996:matrix.org: i do not think if this is easier to setup. | 15:27:45 |
@tiny6996:matrix.org | I figured the only hangup I had on my 3 node setup was setting the ceph storage for host failure instead of device failure | 15:33:31 |
@tiny6996:matrix.org | but it wasn't that hard | 15:33:37 |
gucky | @tiny6996:matrix.org: yes you i know. This was 1 argument to go away of ceph. We tried block storage. | 15:38:14 |
@tiny6996:matrix.org | ceph is kind of weird to get used to. It does everything well but you have to learn to speak its language. | 15:39:12 |
@tiny6996:matrix.org | once you start learning how to talk to it does all sorts of neat tricks | 15:39:27 |
@tiny6996:matrix.org | gucky: what storage system are you using for proxmox? | 15:42:20 |
gucky | @tiny6996:matrix.org: i only local storage. But with an ibm bladecenter you can a failover realize. This has an integrated storage. My old factory had so a setup. | 15:48:16 |
@tiny6996:matrix.org | gotcha so the whole blade will failover instead of just the virtualVMs? that is cool sounds way pricer than what I did though | 15:56:34 |
gucky | @tiny6996:matrix.org: no. The vm from ohne blade card to another | 16:00:25 |
@tiny6996:matrix.org | oh nice I didn't know that worked with just local storage I thought you had to have a separate SAN or cpeh setup for that. | 16:01:20 |