Suspicious incremental backup size

[matteu]

Hello,

I would like your recommendation about this alarm.
My customer has a lot of "false positive" on this on a lot of VM and not necessary DB ...
DHCP server have often arround 500MB incremental size but sometimes 3GB.
Is there a way to easely be able to keep this alarm or not ? The idea is really good, but I have arround 80% of my VM with this alarm several times per week. I just would like to know what advices can you give me to use it or do I just need to disable it because there is no really "good" solution. It's "useless" if I receive like 20 false positive per day.

Does improve Analysis depth to 14 or more can help me to avoid false positive ?

[RomanK]

Hello matteu,

It is similar to the unusual job duration. Simple math takes the last X values, the average, and then compares the outcome with the thresholds.

What is a real suspicious size in that case? Averaging 14 points most probably makes the alarm useless. Again, the same case as with the job duration alarm - some experiments are required.

For example, with analysis depth 3 there are: (500+500+3000)/3 = ~1300
would be
warning 150% = 1950
warning 200% = 2500

Expecting numbers around 3000 I'd set
warming 250% (3250) and error 300% (3900)

Something like that.

Thanks

[matteu]

Thanks for your answer and explanation.
I understand better how it works now

[vmtech123]

Can you adjust this per VM, or backup? if not can that be added in the future?

[RomanK]

Hello vmtech123,

Adjusting per VM is not possible. It can be adjusted per VBR servers, there are also include and exclude filters by Job name.
Would it work for you?

Thanks