Kunal,

There could be many factors.  The CAS service typically depends on some kind of 
back end for authentication and attribute release.  And you probably have a 
load balancer in front of the service and are running multiple nodes.  I would 
start by pulling my logs for the various dependant services and measuring how 
quickly they are receiving and responding to requests.  E.g. if I am using LDAP 
as an attribute store and MySQL as a ticket store, I'd look at those logs and 
see how fast queries were coming in, and how long it took for those requests to 
be filled.  CAS won't be able to complete certain requests without responses 
from those dependant pieces.

You could also try swapping out various components.  E.g. do you get much 
better response times if you use an in-memory ticket store with a static JSON 
attribute store?  If so, maybe tuning those components is the answer.

You didn't mention the CPU frequency you were using-- if the bottleneck is 
CPU-bound, that could be something to look at.

Also, it's worth mentioning, we are measuring somewhat different things.  In my 
test, I am only interested in 6 specific authentication events (TGT 
create/destroy, Authentication success/fail, ST create/validate).  This isn't 
exactly the same as performance testing.

Thanks,
Carl Waldbieser
ITS Identity Management
Lafayette College

----- Original Message -----
From: "Kunal" <[email protected]>
To: "cas-user" <[email protected]>
Sent: Wednesday, October 25, 2017 9:36:33 AM
Subject: Re: [cas-user] CAS Performance Bench-Marking

Hi Carl,

Thanks for the quick response! Appreciate the details! I have gone through 
this.

Actually, we ran our test with 200 concurrent login (per second) having 
server configuration of 4 cores and 8 GB of RAM. The average time taken by 
each request was 4.7 seconds, which is at much higher side, we want this 
time up to 1 second.

We are not able to understand, is this the idle response of CAS server OR 
are we missing something and what's should be the expected response time 
for this server configuration.

Thanks,
Kunal




On Wednesday, 25 October 2017 18:37:08 UTC+5:30, waldbiec wrote:
>
> Kunal, 
>
> There are really a lot of variables to such a test. 
> I've performed load tests for CAS 5.0.x [1] and CAS 5.1.x [2], and I tried 
> to include good summaries of my observations.  Much of the load that your 
> CAS can carry is going to depend on your deployment architechure.  There is 
> a section of the CAS 5.1.x manual that describes how to conduct your own 
> testing [3]. 
>
> Thanks, 
> Carl Waldbieser 
> ITS Identity Management 
> Lafayette College 
>
> [1] https://apereo.github.io/2017/06/09/cas5-perfresults-LafayetteCollege/ 
> [2] 
> https://apereo.github.io/2017/09/25/cas51-perfresults-LafayetteCollege/ 
> [3] 
> https://apereo.github.io/cas/5.1.x/planning/High-Availability-Performance-Testing.html
>  
>
> ----- Original Message ----- 
> From: "Kunal" <[email protected] <javascript:>> 
> To: "cas-user" <[email protected] <javascript:>> 
> Sent: Wednesday, October 25, 2017 7:14:32 AM 
> Subject: [cas-user] CAS Performance Bench-Marking 
>
> Can you please let me know what's the performance bench-marking is 
> provided 
> by CAS in terms of concurrent load CAS can support for given server 
> configuration. I am using cas 5.0.3. 
>
> Thanks in advance! 
>
> -- 
> - Website: https://apereo.github.io/cas 
> - Gitter Chatroom: https://gitter.im/apereo/cas 
> - List Guidelines: https://goo.gl/1VRrw7 
> - Contributions: https://goo.gl/mh7qDG 
> --- 
> You received this message because you are subscribed to the Google Groups 
> "CAS Community" group. 
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected] <javascript:>. 
> To view this discussion on the web visit 
> https://groups.google.com/a/apereo.org/d/msgid/cas-user/fb8fe0c6-2cb3-4bca-898e-b495dc160e7a%40apereo.org.
>  
>
>

-- 
- Website: https://apereo.github.io/cas
- Gitter Chatroom: https://gitter.im/apereo/cas
- List Guidelines: https://goo.gl/1VRrw7
- Contributions: https://goo.gl/mh7qDG
--- 
You received this message because you are subscribed to the Google Groups "CAS 
Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/apereo.org/d/msgid/cas-user/e615f2b3-020f-4397-bca4-9e8c9ea302d0%40apereo.org.

-- 
- Website: https://apereo.github.io/cas
- Gitter Chatroom: https://gitter.im/apereo/cas
- List Guidelines: https://goo.gl/1VRrw7
- Contributions: https://goo.gl/mh7qDG
--- 
You received this message because you are subscribed to the Google Groups "CAS 
Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/apereo.org/d/msgid/cas-user/1693783559.14117290.1508940615729.JavaMail.zimbra%40lafayette.edu.

Reply via email to