Veeam v8 Update Impressions

[LarsWulf]

Hi everybody,

I just updated our Veeam installation from v7 to v8 and wanted to share my experiences with you. First of all the update of the Veeam Server, Enterprise Server and all 5 Proxies was successful without any error. But there are a two things I wanted to share that might be usefull for other people how plan their v8 update:

1) The Veeam Proxy Service got renamed. It is now named "Veeam Data Mover Service" so for everybody who is monitoring the Service state and is filtering by Display name will notice this. It is not an issue at all, you just can change the monitoring and everything is fine again.
Just one question to Veeam: Why are you disabling the old Veeam Backup and Indexing Servive instead of removing them ?

2) Updating the Enterprise Manager DB consumes quite a lot of disk space. In our case we had a 5 GB Enterprise Manager DB with v7 (no big logfile as we use simple recovery option for the DB). Because of the update of the Enterprise Manager the DB file itself is now 7.7 GB and in addition the log file ended at 9.5 GB. So I needed 3,5 times diskspace for my Enterprise Manager DB during update process. This nearly got a problem as the Enterprise Manager runs on a physical host with only a 73 GB RAID 1. If I haven`t cleaned up the Veeam log files during installation the Enterprise Manager DB update would have filled up my complete disk.

I will keep you updated about the results of the first backups running tonight.

[Vitaliy S.]

Hi Lars,

In order to be protected from service change name, I would recommend using "service name" instead of "display name" in your monitoring tool.

By the way, here's another similar discussion on v8 update impressions

Thanks!

[foggy]

Just one question to Veeam: Why are you disabling the old Veeam Backup and Indexing Servive instead of removing them ?

Have you tried to reboot the backup server? Do you still see old services after reboot?

[LarsWulf]

Hi Lars,

In order to be protected from service change name, I would recommend using "service name" instead of "display name" in your monitoring tool.

Thanks!

Hi Vitaly, you are absolutely right. At least during this update I learned this lesson.

Have you tried to reboot the backup server? Do you still see old services after reboot?

Yes all servers got rebooted once after the updated before I looked at the services. And they are still existing with disabled state. To be 100 % sure I just rebootet one of my proxies but still the old sevrices are there with disabled state.

And now to the experience of the first backup window with v8. I have to say v8 is definitely faster. Currently I can`t really point out any numbers because the performance increase has been too much for my target storage. The increased speed ended in higher storage load that caused several VM backups failing with:

Code: Select all

Processing XXX Error: Insufficient system resources exist to complete the requested service. Failed to read data from the file [XXX2014-12-03T214525.vrb]. Agent failed to process method {Stg.IsObjectExist}. 

So for tonight I decreased the limit of concurrent tasks per Repository and hope that this will bring only successful jobs again. I will keep you updated about the results and the performance.

[veremin]

The said issue might have something to do with highly fragmented backup file; might be worth checking. Thanks.

[foggy]

Yes all servers got rebooted once after the updated before I looked at the services. And they are still existing with disabled state. To be 100 % sure I just rebootet one of my proxies but still the old sevrices are there with disabled state.

I recommend to pay our technical support team's attention to this.

[LarsWulf]

Backups are still failing. Even after reducing the concurrent jobs count by 50%. I now created a support call to see what is causing this issue. (case #00701165)

I will keep you updated and share my results.
I also added a hint regarding the disabled and not removed old services to the call.

[LarsWulf]

Just to keep this up to date. My backup is now working with the same speed as before.

I got a private fix for BUG IDs: 39914 and 38623

But these will be part of the Patch 1 anyway that will be released within hours I hope.

Overall I can say v8 is very nice.
Now I will work on implementing the new features.
First of all I will switch my Reversed Incremental and my Incremental with Synthetic Fulls to Forward Incremental forever what will give me hopefully reduced storage load compared to the reversed Incrementals. And the Repository which had Synthetic Fulls will have only one full backup instead of 5. So I will have much lower space consumption so I can use these Repository for more jobs in future.

Second I am exited to role out the Self Service Portal to our users. I will keep you updated about my impressions and ideas how to improve.