Some performance measures

[Pavel Ausianik <Pa...@epam.com>]

Hello,

I have made some performance comparisons between soap 2.3.1 and my current
code (including all my patches proposed), replacing separately client and
server libs

Here are results:

Client 2.3.1, Server 2.3.1
100.842
96.701

Client current Server 2.3.1  
92.733
89.186

Client 2.3.1 Server current
61.186
57.999

Client current, server current
52.249
49.921


As you see Server applications benefit most - up to 40-45% to 2.3.1 code, or
better say -40 sec on my machine on processing 4000 calls. It's hard to
count % in real life, since  it very depends on what configuration  client /
server has.
Client achievements are less significally  - up 10-15% to current code,  or
-8-9 sec. Probably client code was less affected...

Best regards,
Pavel

Re: Some performance measures

[WJCarpenter <bi...@carpenter.org>]

> Client achievements are less significally  - up 10-15% to current code,
>  or -8-9 sec. Probably client code was less affected...

This is consistent with informal client-side comparisons I've done
along the same lines.  The last I did was with the nightly Apache
SOAP from 2002-11-13.  We saw a few percent improvement for some cases,
but we actually had a few cases where it was a few percent worse.
(We have a case with 5 MB of tiny RPC SOAP-encoded response with lots
of tiny elements.  We have another case of a 2 MB RPC SOAP-encoded
response consisting of a single huge string wrapped inside CDATA.)

Keep up the good detective work, though!  It's still much easier for
us (and I presume lots of other people) to move to a newer Apache SOAP
than to migrate to Apache Axis for existing projects.




--
To unsubscribe, e-mail:   <ma...@xml.apache.org>
For additional commands, e-mail: <ma...@xml.apache.org>

Re: Some performance measures

[WJCarpenter <bi...@carpenter.org>]

> Client achievements are less significally  - up 10-15% to current code,
>  or -8-9 sec. Probably client code was less affected...

This is consistent with informal client-side comparisons I've done
along the same lines.  The last I did was with the nightly Apache
SOAP from 2002-11-13.  We saw a few percent improvement for some cases,
but we actually had a few cases where it was a few percent worse.
(We have a case with 5 MB of tiny RPC SOAP-encoded response with lots
of tiny elements.  We have another case of a 2 MB RPC SOAP-encoded
response consisting of a single huge string wrapped inside CDATA.)

Keep up the good detective work, though!  It's still much easier for
us (and I presume lots of other people) to move to a newer Apache SOAP
than to migrate to Apache Axis for existing projects.