moving session from inproc to state server: performance?

  • Thread starter Thread starter Steven Spits
  • Start date Start date
S

Steven Spits

Hi,

We're thinking of moving session from inproc to a state server. My question
is: should we expect a difference in performance? Is it noticeable?

Steven

- - -
 
True,

but I'll take a 15% performance hit any day,
against the possibility that *100%* of my
clients lose their session variables.

Site performance should be measured before making a
move from InProc to State Server Sessionstate management.

If more hardware is needed, to make up for the 15%
performance loss, it sure would be a good investment.

But maybe the 15% perf loss wouldn't matter,
if there's enough of a perf "cushion" which could
serve to compensate that 15% perf loss.




Juan T. Llibre
ASP.NET MVP
http://asp.net.do/foros/
Foros de ASP.NET en Español
=====================
 
I think that 15% figure comes from Dino Esposito's ASP.NET book. My own
tests have shown that it is much less, under 10% for instance on a
consistent basis. There are certainly overriding factors such connectivity
etc. Anyway, 15% is manageable for most real world applications that concern
themselves with reading files and parsing external resources.

--
Regards,
Alvin Bruney [MVP ASP.NET]

[Shameless Author plug]
The Microsoft Office Web Components Black Book with .NET
Now Available @ http://www.lulu.com/owc
 
not particularly. not sure why anyone would use inproc on a production site.
you biggest difficulty will be that objects stored in a out of process
session must be serializable.

-- bruce (sqlwork.com)


| Hi,
|
| We're thinking of moving session from inproc to a state server. My
question
| is: should we expect a difference in performance? Is it noticeable?
|
| Steven
|
| - - -
|
|
 
Alvin Bruney said:
I think that 15% figure comes from Dino Esposito's ASP.NET book. My own
tests have shown that it is much less, under 10% for instance on a
consistent basis.
Click to expand...

Thank you all!

Steven

- - -
 
How do you measure performance?
If you measure it in average response time under light load then you would
be correct.
However, if you're talking about scalability then State Server (And
especially Sql Server) storage can give far better performance under
demanding conditions.
 
Back
Top