My company currently uses Backup Exec 10d without CPS for our backups. Right now our rotation scheme seems very inefficient and I'm posting it for general comments and suggestions as to what we're doing wrong, or what could be done better.
Device:
- Dell 132T, 22 LTO2 Tape slots (200GB/400GB)
- Partitioned into 3 slots
- 17 slot - incrementals
- 4 slot - full backups
- 1 slot - cleaning or emergency slot
We currently have 6 jobs that run:
1. Incremental (every weekday except wednesday)
- Backs up file server and Exchange information store
- usually around 45 GB
- backs up to pool of tapes in 17 tape partition (rotation of 3 pools)
2. Full backup (every wednesday)
- Same resources as incremental
- Backs up to 4 tape partition, tapes swapped out every wednesday morning
- 4 week rotation
- week 1 tapes taken off site for 1 year
- weeks 2 - 4 tapes re-used every month
- Current at 600 GB
3. Two Data-heavy department workstations
- Backs up to 17 Tape pool, incremental and full (alternating nights)
- Each full backup is about 400 GB, incrementals are 20 GB
The problem we're having and the reason for this post is that after the pool gets near full, the larger jobs keep overwriting tapes. From reading these forums I realized its because they are all append jobs, and append jobs spanning tapes overwrite the second tape. I tried setting it so that if it didn't fit, the job would fail, so that I could change the pool (we have a 3 pool rotation) but the job doesn't fail, it just keeps asking for empty tapes.
In the very near future the data-heavy departments are going to severly increase in size, as we will begin backing up data that is currently held on external drives. We want to have a solid schedule in place before this happens.
Any suggestions would be appreciated.
