I always used multiplexing (Oracle side) and didn't see much of effects with and without (there are effects, but ignorable in my case). 1:1 is strange for sure. I also have lot of logs, but still manage to get overall 10x factor. While I was in testing phase from what I could see logs would give you 2-3x, DB 5-6x factors. We do not use compression on pure Oracle systems (we do with SAP systems, but the impact there is different due to internal structure). Our DBAs did test and concluded that some DBs would even grow with compression and didn't want to have mixed bag so they discarded whole concept of compression on pure Oracle databases. Bare in mind that there are two type of compression you can use and if I remember one is friendly for de-dupe while the other is not.
You didn't say how you use, but if VTL create separate pools for DB and LOG and check what compression rate do you get on volumes.
ble1
4 Operator
•
14.4K Posts
1
June 3rd, 2014 14:00
I always used multiplexing (Oracle side) and didn't see much of effects with and without (there are effects, but ignorable in my case). 1:1 is strange for sure. I also have lot of logs, but still manage to get overall 10x factor. While I was in testing phase from what I could see logs would give you 2-3x, DB 5-6x factors. We do not use compression on pure Oracle systems (we do with SAP systems, but the impact there is different due to internal structure). Our DBAs did test and concluded that some DBs would even grow with compression and didn't want to have mixed bag so they discarded whole concept of compression on pure Oracle databases. Bare in mind that there are two type of compression you can use and if I remember one is friendly for de-dupe while the other is not.
You didn't say how you use, but if VTL create separate pools for DB and LOG and check what compression rate do you get on volumes.