I have 2 identical databases. I have page compressed all tables and indexes in one of them. This gave me around a 40-50% saving in terms of storage.
Now when I run the same query (select only) against the 2 DBs I see that the run time on the compressed DB is slower even though the amount of reads is about 50% as compared to the uncompressed DB.
I have measured SQL performance monitor counters (PLE, lazy writer, disk read/write times) and they mostly show that compression had a positive impact. The interesting thing is that CPU usage only increased marginally (1%) – I was expecting a higher increase and of course as I say the query runtime on the compressed DB is about 5-10% higher.
I’m still investigating but any pointers here would be helpful.
thanks
Ishfaq