Lately, a lot of queries to the database are timing out. While investigating I found that queries are in the suspended state with wait_type as PageIOLatch_Sh. I have not noticed so many queries being suspended before on the server. This has been happening for the since Saturday.
A couple of weeks ago, the cpu on the db server was at 100% constantly. While trying to fix that I changed a couple of indexes which were not really being used and were resulting in user_scans. After changing the indexes, the cpu usage on the server went down, the performance of the site was generally faster and looking at the index stats now a high number of user_seeks were happening, which I thought was a good sign.
However, now after a week or two, even though the cpu usage is not at 100, a large number of queries have started timing out and PageIoLatch_Sh waits are being observed. The interesting thing is that these timeouts were not happening at such a high frequency earlier even though the cpu was at 100%.
Now maintaining indexes is also something which slows down processing of sql server, so I was wondering whether the new indexes could have resulted in a larger number of these waits and timeouts?