Task Slots vs. Write Streams on Backup Repositories

[pinkerton]

Dear Forums,

page 45 of the best practice guide for Backup & Replication states that at the repository, which writes the backup data, only one thread is writing to the backup storage per running job. Does this mean that parallel processing of virtual machines within a backup job is effectively pointless unless writing to a repository that is enabled for per vm chain and therefore supports multiple streams? What exactly is the distinction between a task and a write stream?

What I'm currently not understanding either is whether a task slot in the repository corresponds to one proxy task or to one write stream. Let's use the following example:

- One backup job with 2 VMs (one disk each) <-- consumes two proxy task slots
- Second backup bob with 1 VM (one disk) <-- consumes one proxy task slot
- Backup proxy with 3 Slots
- Repository with 2 Slots

Will Backup Job 1 consume both task slots from the repository (which then would mean they equal to proxy task slots) or only one (which means they correspond to write streams)?

Thanks,
Michael

[dellock6]

Hi Mike,
parallel processing in the general options refers to the parallelism of proxies, not repositories. So the two parameters don't count one against the other, and you have to deal with both. Said differently, one repository, even without per-vm chains, will write on a backup file the data coming from multiple proxies that are working with parallel processing enabled.