Search results

As the title says, can you make an option for lxc to use ssh keys instead of passwords? I realize you can set this manually, but i like to keep things automated. or, workaround (easier than dab): untar a template (with 'p' to preserve permissions), add your ssh key, and add "/usr/bin/passwd...

from t.lamprechts earlier answer, i just made this one of my post install scripts (first two lines commented out for production). works fine after running it. will try again without after the next iso release. the mirror or whatever it is should work itself out by then. #!/usr/bin/env bash rm...

That would be strange, but possible. did you try "apt-get update" or "apt-get install unzip" from the default install?

yes, that works. but, this problem is in the default install with proxmox-ve_4.0-0d8559d0-17.iso. heres a packer template you can use to reproduce the issue. https://gist.github.com/xahare/116034564973ab7648cb if you try it, you might want to comment out the provisioners. they are for os x.

after installing proxmox4 to vm for testing, removed the enterprise repo, and added the no subscription one. first apt-get update gets a gpg error. and, i cant install unzip, which is in the main debian repository, https://packages.debian.org/jessie/unzip. also tried parted. is there something...

https://gist.github.com/xahare/116034564973ab7648cb have fun.

Can someone well versed in the pve cluster stack please write a cluster troubleshooting guide? in http://forum.proxmox.com/threads/16814-RGManager-won-t-start?p=86859#post86859 maybe, by default this should start with "<rm/>" when i did my test, i had to start rgmanager by hand. (ansible...

Dockers got you covered. https://docs.docker.com/machine/ just use the generic driver. or make a proxmox one. i would not complain :)

Im sure this is covered somewhere, i just cant find it. How do you make a quorum only node in a cluster that you intend to upgrade to 4.0 (so not using a quorum disk)? in other words, a node thats only there to maintain quorum and wont run any instances. in my case, i have 3 real nodes, and id...

ceph will let you do this. http://ceph.com/docs/master/rados/operations/cache-tiering

mtu is 1500 on both, and yes, same network. its an inhost vmware network on my laptop. but i tried changing the monhost like to space like yours and it worked :D the cluster at work has ';' between the monhosts, so i guess ; and space work, but not ',' root@c1:~# netstat -an | grep 6789 |...

root@proxbox:~# rados -p rbd ls 2015-07-30 12:32:43.067025 7f298e15e760 -1 did not load config file, using default settings. no monitors specified to connect to. couldn't connect to cluster! error -2 root@proxbox:~# grep -A 5 rbd /etc/pve/storage.cfg rbd: rbd monhost...

we have a real cluster (3.4 supported version with kernel 3.10) and ubuntu 14.04 / ceph giant where this stuff works. test "cluster" is single proxmox 3.4 (kernel 3.10) node connected to 3 node ceph cluster, all are vmware fusion virtual machines. guests on them run fine with local storage...

We have 3 nodes of proxmox 3.4 under support and a separate ceph cluster running in ubuntu 14.04. would like to update to hammer on the server. will the clients be ok with the pve repo or should we add the wheezy ceph repo to the pve nodes?

has anyone run proxmox IN lxc? i was thinking of trying that for testing as it would reduce one layer of virtualization if it works.

theres little point to HA without HA on the storage side. since your only using kvm, ceph is your best bet.

microsoft wants a separate license for each possible physical node a vm can run on. the software is licensed per physical computer. is there a way to limit the failover node so we only pay twice as much for windows server licences instead N times as much?

whats wrong with the ceph packages? do the jessie ones not work? [edit] just saw the lack of them http://ceph.com/docs/v0.69/install/debian/

unicast corosync makes testing in virtual machines easier and allows for small networks without fancy switches. note to anyone else reading this: for testing in linux you can "echo 1 > /sys/devices/virtual/net/br0/bridge/multicast_querier"

according to corosync docs, you can use it without multicast. but then you have to give it a list of all the other nodes.