sexta-feira, 11 de maio de 2012 14:59
I am experiencing what seems to be a possible caching configuration issue on our farm. We have developed a WCF service (please don't move to dev forum) that is deployed in another IIS site on one of our WFE servers and this service uses the same application pool ID as the SharePoint Web app to allow content access.
One of the service methods queries a document library and returns a list of document names. This runs blistering fast on the dev farm but, what seems to be sporadic, runs very slow on the pre-prod farm. The service method code is extremely basic so it's my belief that coding isn't the issue.
I thought that configuring the SuperReader and SuperUser accounts would assist with this issue believing that it could assist with data caching and then started noticing that with consecutive calls, the service speeds up and delivers the result in seconds where else, it takes almost 40 seconds. This burst of speed doesn't last forever and later in the day, our slow response persisted.
There are currently only 5 documents in the library so, again, volumes are not an issue either.
The speeding up of the WCF service call to the library leads me to belief that the data is getting cached somehow. I am uncertain as to where this is cached or how to adjust it.
In addition, I've made ensured that nothing else runs on the server farm and also concluded that the servers' CPU, memory and disk operations are actually nominal during the testing.
Can someone please explain to me what this behavior could be? Am I on the right track about it possible be caching? Any additional troubleshooting tips?
Your input on this is greatly appreciated
Todas as Respostas
sexta-feira, 11 de maio de 2012 19:29
Not a caching issue.
Simple questions first, did you make sure that the pre-prod farm is not calling the dev farm?
sexta-feira, 11 de maio de 2012 20:38
Yes, the 2 farms are on a seperate domains and ipranges altogether. No comms possible between them.
segunda-feira, 14 de maio de 2012 16:59one possibility is that the authentication for the WCF call is taking long. are you passing in user/pass when calling or using the app pool account? Also i assume the servers have plenty of memory and processors. none maxing out and slowing things down.
- Marcado como Resposta Rock Wang– MSFT quarta-feira, 6 de junho de 2012 08:54