Search results

Try assign 2 cpu's to the machine. I recall this as being a known bug - and that is the fix. Sent from my iPhone using Tapatalk

I suggest... 802.3ad And jumbo frames. Should be a bit better :)

What type of bonding have you used?

Are you using NFS? or ISCSI? And how have you made that caluclation of speed? The only system I know that gets near wire speed is good ol' ftp. And even that would not be 250MB/s Remember.. complex protocols. (SMB,SSH,NFS,ISCSI,etc) have lots of hand shaking. End result is you loose...

Sorry... That is exactly what I meant. I should read what I write more carefully before submitting it! Sent from my iPhone using Tapatalk

I don't think you read the bit about formulating complete sentences? Your bullet point ? Lists are impossible to work out! Can you try actually explain what you are asking? Sent from my iPhone using Tapatalk

Your limit is cased by the network layer. In my experience... 1gb links are not fast enough. Local disk: 300gbs 1gb link: 90gbs if you are lucky! So.... Bonding helps. 2 cards will give you 150gps or close to that. If you want speed.... Fibre channel is really the only answer :)...

I sadly have little choice but to use RC1 for production as of this weekend. A new cluster has to go live - and I don't fancy having to upgrade from 1.8 later! Sent from my iPhone using Tapatalk

If I am correct.. the Sil range of raid cards are not really true raid cards. They have never been well supported in linux. Rob

If you are running a 'unix' system I would recommend a container based approach. Check this link: http://wiki.openvz.org/Physical_to_container For a good example of how to do it!

Add the following too... /etc/sysconfig/network GATEWAY=XXX.XXX.XXX.XXX This option should be created but seems to be missing on the container templates. Should sort your problem :-)

You will need nfs or iscsi to live migrate. Does your shared storage support Nfs? Sent from my iPhone using Tapatalk

Hi All, Just wondering what the rough eta is for proxmox 2.0 to be released? I have a new cluster due to go live shortly - but have rather cautious about using the beta on a live environment! Lastly.. assuming I use 1.9. Then will there be an upgrade possibility to 2.0? Many thanks Rob

Hello, I have an odd issue with proxmox 1.9 I have started to migrate my machines to it - and now find the guest machines loosing their network config. All my OPENVZ Guests have an eth0 interface mapped to them. This is necessary as each interface. e.g. VMBR7 is on its own vlan. So.. we...

You could always do a: find /var/lib/vz -name "*id*" Or... Just make sure the cron for mlocate runs at a time that does not interfere! Rob Sent from my iPhone using Tapatalk

This may have come up before.. but I thought worth mentioning. If you occasionally have backups using the vzdump system fail - because the /mnt/pve/vzsnap folder could not be unmounted...The issue is most likely the mlocate cron job traversing this mount point.The fix is very simple - and...

Personally.... There are two I recommends. 1. Vyatta 2. PfSense Both are free - and can also run in a vm guest if required. Both are very different and suit different types of users! Rob Sent from my iPhone using Tapatalk

I have always approached vm machines on proxmox by using a bridged interface, then configuring the machines eth0 port. Doing this makes the system much closer to a real machine for day to day networking and configuration. The port forwarding then can just be dealt with on the router! Sent...

My scripts are attached. All they do is to create the lvm volume as /var/lib/vz.snapshot and remove it. Simply modify to suit, and configure bacula to execute them on start and stop.. and to ignore /var/lib/vz Rob

Hello, I am rather paranoid! I do the following: machine internals backuped up externally to livedrive.com using duplicity lvm snapshot and backup of /var/lib/vz weekly vzdump Somehow it all works! Rob