- Posts: 264
- Liked: 30 times
- Joined: Mar 22, 2011 7:43 pm
- Full Name: Ted
1) WAN bandwidth saturation.
a) During any pre/post processing (validating task, rollback-on-failure, snapshot creation/removal) you can still be using your WAN.
b) More traffic for any WAN optimization software to keep saturated.
2) Chaining without waiting. If I get a replication job that wants to run for 48 hours, then nothing else will run until it's finished. I think I could create a complex powershell script, but I also think it would be good as part of the product. Example:
8 VM's -- 1 through 8 in order of priority.
I want 2 VM's processing at the same time, continually looping 1-8 (or starting once/day or whatever schedule).
Start with 1 and 2 running. When either one finished, continue with 3. When either one of 3 or (1 or 2) finishes, then start 4, etc.
I originally handled this by creating 2 separate manual chains. 8 separate jobs -- 1 calls 3 which calls 5 which calls 7. 2-4-6-8. The problem is if 3 happens to have a lot of changes, then 5 and 7 might not run for days waiting for 3 to finish.
Veeam 6 may help this, as replications (pull configuration) won't be limited to 3 simultaneous jobs before you run out of connections to the destination host.
Maybe this is a good summary way to say it -- If a virtual machines is going to take a long time (2 days for example) to replicate anyway, I would prefer to run additional virtual machine replications in parallel even though it will make the 2 day replication job take even longer.
- SVP, Product Management
- Posts: 29341
- Liked: 5470 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland