Is there anything I can do for a OLEDB wait type?
After restarting my instance, I started a DBCC checkdb on a 200 GB database that uses the spatial data type. A couple of tables have 250 million rows. There are no indexes on computed columns or filtered indexes. It's been running about 2 hours and the percent complete is about 2%. The estimated completion date (4 days from now) has been increasing by about and hour a minute. It might be weeks before it completes. (I'm using a db restored on my desktop - 6-core Zeon, 16 GB, and 2 SSDs - with SQL 2012 Developer SP2. I enabled trace flags 2562 and 2549 enabled. Compatibility on the db was changed to 2012 from 2008.)
I looked at the wait stats and see the OLEDB wait type at almost 100%. When I look at the activity monitor "Resource Waits", the only wait I see is Network IO - about 1 sec per sec. "Data File IO" is at 0.0 MB/sec with response times normally at 0 ms, but can be 2 ms from time to time. CPU use is about 10%. My machine feels idle.
An AVG scan recently started and just finish. It did not seem to slow anything down or change the waits. But that makes me wonder. Can a real-time virus scanner be a source of the OLEDB waits? Would it also show up as network io wait?
Any advice that can help me find the bottleneck would be appreciated. It would great if I could move the bottleneck to CPU or disk.
Thanks.
Randy in Marin