Finally run out of ideas on this one....
Just to get it outta the way, the environment:
2 NLB Web servers (each Win2003, IIS6, 2G memory, 2 physical/4 logical
CPU, 2 NICs, Unicast)
1 DB Server (Win2K, SQL2K, 2G memory, 2 physical/logical.. no
hyperthreading)
I'm periodically getting a "Timeout Expired..." error on
SqlCommand.ExecuteReader for VERY simply INSERTs and UPDATEs (no
stored procedures) ex. INSERT INTO MyTable (fields) VALUES(values),
UPDATE MyTable SET... WHERE MyTableKey=@MyKeyValue
It happens on a variety of tables, none of which have triggers. The
SQLCommand timeout value is set to 60. SQL Server is showing no
blocks, long lock wait times, or anything else that would indicate a
problem on the DB server. No backups or replication snapshots at
these time either. In fact, the DB server is humming along nicely.
Also, using non-pooled connections or pooled connections does not seem
to make a difference.
There are some strange things:
First, only seeing it at the top of the hour, but there is nothing
specifically running at the top of the hour I can find. (no jobs,
recycling of app pool, other services running) Doesn't seem to be
related to load (sometimes happens at peak, other times off peak (ex.
8pm)
Second, only happens on sites that are NLB... there is an asp.net
site on one of the servers (same DB, similar C# code) which does NOT
use NLB and does not see the errors.
The only thing I can see that COULD be a problem is that the DB
Server's Subnet is the same as Web Servers Virtual and Dedicated IPs
on NIC #1 (versus our other subnet on Web Servers' NIC #2) I'd love
to blame this, but doesn't explain the consistent time of day when I
see this.
Any ideas?
Thanks,
-- JD
