Unsolved
This post is more than 5 years old
16 Posts
0
1512
Avamar VADP backups failing with Timed Out -Start error
Hi All,
We have installed an Avamar 6.1 gen4 single node integrated with DD for the customers VADP backups. we also have a COLO where we have the same setup of Avamar and Data domain and we hvae one way replication between the two sites.
After a few successfull backup/restore tests with 3 servers we added 20 VMs for backup at once and they all failed with this error. we have close to 100 VMs and only one proxy server which was deicided with the help of the sizing tool. I checked for reverse lookups from the Avamar server to the client and they seem to resolve just fine. But I was not able to telnet to the clients 28002 port from the Avamar server. I am not sure if that is relevant here as there is a proxy in between.
Neverthless I have been unable to find a solution to this and would like to have some troubleshooting steps that I can go with. Also the logs from the Activity monitor are blank.
Do we need to add more Proxy servers to handle so many VMs?? Also is there any other tuning or additional steps like checking for ports blocked by the firewall that need to be opened up??
Thanks in advance for the help!!!
fdxpilot
143 Posts
1
December 13th, 2012 10:00
You're likely to need more than one proxy to backup your VMs (especially for the first backup of each VM which can take a while). Was the proxy busy running backups for the entire backup window? If so, the jobs remaining to run will end with "Timed out -- start" status when the backup windows ends.
KeyurD
16 Posts
0
December 13th, 2012 11:00
ok, is it possible to take those proxies out once we get the first full backups done?? We are doing a once weekly backup of the VMs with CBT enabled.....the customer doesnt want to spin more VM proxies because of Storage space constraints.
thanks.
fdxpilot
143 Posts
0
December 13th, 2012 12:00
Yes, it may be possible to reduce the number of proxies you use once all of your VMs have a backup seeded to Avamar. This is especially the case if your VMs have changed block tracking enabled. Generally, this will reduce the amount of time required to backup a VM. However, if you are only running one backup per week there may be a lot of changed data to backup each time. This could cause you to need more proxies since each backup job will run longer. In addition, if you have a large amount of changed data each time a backup runs, the effectiveness of CBT will be reduced. Trial and error is the best way to figure this out. Each proxy requires a relatively small amount of disk space.
I would added new VMs gradually and see how many proxies are required once each VM has a backup completed to Avamar.