Search results

Yes, the 3.10 kernel is still experimental for this very reason. If youre feeling adventurous and know the ins and outs of openvz: https://openvz.org/Vzctl_for_upstream_kernel

Yea, if you buy the community support, you get just that: community support. Meaning: the non-subscription repository gets all updates first (after they pass internal testing) and then the community can use and test them with the tons of different hardware combinations out there. A little while...

first of all it is strongly suggested to NEVER EVER change the IP address of ceph nodes (more specifically: MONs). Ive you absolutely have to, you can try this: http://ceph.com/docs/master/rados/operations/add-or-rm-mons/#changing-a-monitor-s-ip-address

neato. I wonder whether thered be interest in translating this to german and how complicated that would be lector/publisher-wise.

as far as I know the openvz guys are currently developing CRIU to be able to handle snapshotting not only for containers but for individual processes as well. It's still in development though, so I am merely giving you a heads up. What you can do to at least get filesystem snapshots of openvz...

theres actually a difference. When you create a snapshot, you create a CoW-image of the original image, meaning that any changes to the disk after this point will be saved to the new image, allowing you to go back to the original image (which also becomes read-only ("immutable") when creating a...

doesnt HA automatically move all the VMs back onto the node youre emptying if you were to use something like that?

You don't exactly NEED raid controllers for ceph to function. - You need raid controllers if you want more disks in your system than the mainboards controller is offering - You can use a raid controller to benefit from its battery-backed RW-caches. Do note that regular hard drives already have...

you may want to look at this benchmark: http://ceph.com/community/ceph-bobtail-performance-io-scheduler-comparison/ while it is actually trying to determine which IO scheduler is best for several setups, you can also see that the 8xRAID0 setup (meaning: 8 disks as 8 RAID0s) in almost all...

You don't need 3 virtual server to host 3 websites on 1 IP. you can host all of them with one webserver. Now if you want to separate the sites for security reasons, thats completely valid. Just saying that you dont need to use virtualisation if you don't want to. If you want to, you may want...

Just an FYI: people on the ceph-users mailing list have reported drops in ceph performance after upgrading from dumpling to firefly and from emperor to firefly respectively, with decreases of up to 20% in performance. Investigations for this are apparently underway atm ("Yeah, it's fighting for...

jumbo frames are useful for ceph, just be aware that your switches need to support them.

The only reason why the Ceph guys recommend putting the cluster communication into a separate, private, non-routed network is for security purposes: so that it's impossible for anybody to disrupt the cluster communication by flooding the network. That said, its not actually a necessity, ceph...

hm... weird. Maybe you can try this: get ceph.conf and ceph.client.admin.keyring from /etc/ceph of the ceph cluster (or ceph-deploy) and put it into /etc/ceph on the proxmox node. With that you can use the ceph CLI from the proxmox node and see if that connects okay when you turn off .2 So...

did you enter all the monitor addresses to proxmox? like so: because if your proxmox only knows of 1 monitor and you take that one down for testing... well

This delay is there because a ceph cluster doesnt need the data to be instantly moved when a node goes down because the others will continue to serve the data. The cluster will rebalance data after the timeout has been reached. If your ceph isn't continuing to serve data when one of the nodes...

yea openvswitch is a way to implement SDN (software defined networking). you can either use vlans to connect VMs on different nodes or GRE (switch needs to support MTU > 1500)

the only thing really is that the network hardware connecting your nodes needs to have all the VLANs youre going to use mapped to the trunking ports that connect to the nodes

I think I fixed it, I noticed how for reasons that are beyond me, apt was only listing outdated ceph-common versions (arent firefly packages supposed to be in the debian stable main repository?). I had to add deb http://ceph.com/debian-firefly/ wheezy main and do "aptitude upgrade ceph-common"...

Hi, does the pve-manager dump logs or error messages somewhere, indicating why the ceph connection is failing? In the web interface I can only see a connection problem, which isnt saying much. storage.cfg: rbd: rbd monhost 192.168.178.41:6789;192.168.178.42:6789;192.168.178.43:6789...