I'm currently using EF6a3 in a REST web service (written using WebApi). I've done some simple stress testing using tinyget to hit one of my methods which does some simple database access (via EF) and makes one external REST service call. When I run it under IIS using 25 threads, looping 50 times, it takes ~37 seconds. When running the same test under Katana, that number goes way up to 95.89. I've isolated it to the database layer. If I skip that, the performance number even under Katana drops way down to around 17 seconds.
I'd like to test if the problem is still there under EF5 but downgrading doesn't appear to be very easy. I did change my async EF calls back to synchronous but that didn't seem to make a difference. I'm not really sure where else to go from here. Can anyone think why EF would be so much slower under Katana?
FYI, to run my service under Katana I'm using this for middleware...
http://nuget.org/packages/Microsoft.AspNet.WebApi.Owin
tim
I'd like to test if the problem is still there under EF5 but downgrading doesn't appear to be very easy. I did change my async EF calls back to synchronous but that didn't seem to make a difference. I'm not really sure where else to go from here. Can anyone think why EF would be so much slower under Katana?
FYI, to run my service under Katana I'm using this for middleware...
http://nuget.org/packages/Microsoft.AspNet.WebApi.Owin
tim