SSL Performance HIT in REVERSE? the clients are PEGGED as opposed to the server?

[anon]

I am trying to stress test the effects of SSL on a web server using Web
Application Stress Tool (WAST).

However, when using SSL and according to the WAST Help files, the peformance
HIT is in the opposite direction. It's on the clients at 5 times the load!!!
as opposed the server I want to test. So the help files are correct.

Therefore, I can't even test the server for SSL until I significantly
upgrade my client or clients to even deliver the stress in the first place.

It seems like SSL stress testing is tougher on the "tester" as I THINK, it
needs to do some decryting of the sending page?

So, it's the clients that are doing most of the work on SSL as opposed to
the server?

Can anyone enlighten me more on this subject. As I wanted to see the effects
of SSL on the web server as opposed to not using SSL.

[Ken Schaefer]

Please post only to relevant groups. This has nothing to do with VB.NET (for
example), or C#

Both the client, and the server, are being stressed when using SSL. The
communication in both directions needs to be encrypted at one end, and
decrypted at the other. So, the client makes a request (and encrypts the
request details). The server receives the request (and decrypts it). The
server sends a response (and encrypts the response), and the client receives
the response (and decrypts it). Both of these operations (encryption and
decryption) are CPU intensive.

Cheers
Ken

"anon" <anon@hotmail.com> wrote in message
news:eZtqSLWPEHA.272@TK2MSFTNGP12.phx.gbl...
: I am trying to stress test the effects of SSL on a web server using Web
: Application Stress Tool (WAST).
:
: However, when using SSL and according to the WAST Help files, the
peformance
: HIT is in the opposite direction. It's on the clients at 5 times the
load!!!
: as opposed the server I want to test. So the help files are correct.
:
: Therefore, I can't even test the server for SSL until I significantly
: upgrade my client or clients to even deliver the stress in the first
place.
:
: It seems like SSL stress testing is tougher on the "tester" as I THINK, it
: needs to do some decryting of the sending page?
:
: So, it's the clients that are doing most of the work on SSL as opposed to
: the server?
:
: Can anyone enlighten me more on this subject. As I wanted to see the
effects
: of SSL on the web server as opposed to not using SSL.
:
:

[Cor Ligthert]

Hello Anon,

Can you tell us in which way this question is VB.net language related, maybe
we can give you an answer from that newsgroup?

Cor

[anon]

Of all the newsgroups, VB programmers tend to explain things better than
most other programmers, that's why.



"Cor Ligthert" <notfirstname@planet.nl> wrote in message
news:e0vxqXWPEHA.904@TK2MSFTNGP12.phx.gbl...

> Hello Anon,
>
> Can you tell us in which way this question is VB.net language related,

maybe

> we can give you an answer from that newsgroup?
>
> Cor
>
>

[Keith W. McCammon]

[Thought about removing the long list of irrelevant groups, but it's too
much work...]

> However, when using SSL and according to the WAST Help files, the

peformance

> HIT is in the opposite direction. It's on the clients at 5 times the

load!!!

> as opposed the server I want to test. So the help files are correct.

Is the issuing certificate authority trusted by the clients?

[Keith W. McCammon]

> Of all the newsgroups, VB programmers tend to explain things better than

> most other programmers, that's why.

Heck, if I always wanted a thorough and concise explanation for things, I
wouldn't even bother posting to Usenet half of the time...