RPC Timeout Error

  • We recently moved our servers to an offsite data center. For the last year, my users have used an app I created to import data from an XML document into our local database server. Virtually the only thing running on that server was my database and a couple of other rarely used databases. This was all local and it was fast.

    Our new sql server and network at the data center is blazing fast (supposed to be) with ample memory and hard drive space in one blade. The Sql Server is on a virtual machine and is set up as a server. RAID drives are in use. This sql server contains our Shairpoint database, Team Foundation Server database, my database that I recently moved there, and several other databases. Now, when I run my application from a virtual machine at the data center, I receive periodic RPC timeout errors. When I run it from my workstation, it seems to work fine but is EXTREMELY slow compared to before.

    I have viewed the SQL Profiler while the import is running and each insert doesn't even register as taking any time at all; however, the time between inserts appears slow. Likewise, Sharepoint and Team Foundation Server seem to be hitting the database constantly. The profiler shows sp Reset Connection being called many times.

    Please guide me where to start. Should my app be moved to a different database server than one containing Sharepoint and TFS? Are all of the calls to Sharepoint and TFS causing the problem and are they necessary? Could this be a network issue? I didn't think it was a network issue because when I query the database it is fast. Is there some RPC security going on with the firewall? I don't have hardly any indexes setup on these tables which shouldn't slow the insert down.

    Thank You.

  • this is so difficult to answer .. i have to be diplomatic as we too have an outsourced data centre 🙂

    Experience has shown that in certain circumstances the combination of shared resource and virtual servers has caused severe performance issues. Performance can also be inconstant ( in my case based around the time of day ).

    All I'll say is that what is claimed and the reality may be far apart. e.g. Just because your servers are all connected with gig ethernet doesn't mean you get the performance of gig ethernet - the number of ports used on switches may be high leading to contention - this can also be the case with storage too. trying to work through VM performance is a nightmare with the likely result of complaining about poor performance being a set of graphs which show everyhting is hunky dory on the VM !

    You likely need to attempt to run tests to see if there are bottlenecks, it's a difficult situation for sure. I'd try complaining first off by raising support calls and see what answers are returned.

    I have sharepoint and TFS but they don't show much resource use on my servers, possibly due to low numbers of users. Remote timeouts and such are tricky - I did have issues and some changes were made to firewall settings, nic settings on remote workstations ( e.g. my PC(s) ) and maybe some settings on rdp client.

    [font="Comic Sans MS"]The GrumpyOldDBA[/font]
    www.grumpyolddba.co.uk
    http://sqlblogcasts.com/blogs/grumpyolddba/

Viewing 2 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic. Login to reply