Hi There; I’m an ‘Accidental DBA’ with a problem (is there any other kind?).
We seem to be getting very high ASYNC_NETWORK_IO; with a WaitCount of around 20 million hits per 24 hour period (total wait time around 9 hours in that same period).
I’ve spent a lot of time researching this wait and as I understand it ASYNC_NETWORK_IO is most often caused by the application not consuming data fast enough; so we had a programmer work through the code and resolve all locations where we are using a IQueryable in a for next loop and converting them immediately into arrays (Linq To SQL).
This seems to have made no difference.
I would like to set up an extended event trace to find which queries are generating this particular wait, but at an average of over 200 hits per second I’m concerned about performance impact such a trace might have.
I’m looking for suggestions as to how to move forward in resolving this issue.
My first question is, am I correct in thinking that the number of these waits that we are getting is excessively high?
Assuming that is the case how can I go about tracing the offending queries without hammering the system?
If (as I suspect) it is not individual queries that are causing the issue, what else should I be looking for?
Thanks in advance
Paul.