Server Manager extremely slow

I've been on at least 20 different systems server manager from 8.9x to 9.1 and I've never seen one as slow as my current client. It's extremely bad when looking at the user sessions, call object stats and caches for the JAS servers. Most of the other functionality is not too bad, but still a little slow. It is taking up to 2-3 minutes to load the user sessions and there's only 15-20 users on and taking well over 10 minutes to load the call object stats page. I know about all of the problems around the Server manager console heap out of memory and we are not running into that issue. Anyone ever seen anything like this before?? I'm starting to lean towards something mis-configured in the network but I don't see how it would just make certain parts of it so slow and not others. Also, the JDE web client is performing just fine with regards to performance.

Thanks in advance for any insight. I have a call open with Oracle as well, so will keep this updated if we work to a solution.

System info: JDE 9.1/9.1.0.5 tools all around. Oracle database 11.2.0.3 on AIX 7.1, enterprise server AIX 7.1, Windows 2008 R2 64-bit for deployment server and JAS servers, Weblogic 10.3.5

How are your users connecting to your WebLogic servers? We had our users going through a Barracuda load balancer, and saw similar very slow response times. I think that server manager was trying to resolve IP addresses into a computer name. Originally, we didn't have a DNS entry for our load balancer, so my guess is that it was trying to resolve the IP address, waited for the timeout, and then went onto the next entry. It was very SLOOOOW. Then we created a DNS entry for our load balancer's IP address, and response time picked up tremendously. I don't know if this is your situation, but thought I'd share what our history was with a similar issue.

Interesting...we're not using a load balancer yet, but we will be using an F5 in the future. We are on a segregated network from their normal production one so it very well could be something similar with regards to DNS or some other kinds of WINS resolution. Thanks for the input.

To test this, just try typing in an entry for the deployment server (or wherever you're running server manager) in the workstations HOSTS file. That way, you won't be inquiring against the DNS server constantly - and you should see a performance boost.

Otherwise, it could be antivirus on your deployment server ? Just make sure that the java libraries are excluded from the antivirus scan. Easy way to check to see if this is the issue is to monitor task manager as you access server manager to see what processes are maxing out - if its antivirus, then you'll see it.

Lastly, check to see what Java SDK you're running. From the command prompt, run "java -version". Check to see if the java SDK is being specifically referenced in the server manager startup. It _might_ be the cause - but its usually not.

We are experiencing the exact same thing as you are reporting. We first noticed this when we upgraded tools release from 8.98.4.4 to 8.98.4.7.

I opened a ticket with Oracle and we tried several things. We ended up upgrading just the Server Manager to 9.1 tools. This helped with overall performance of server manager but the user sessions, database caches, etc. are all still slow. The more users on the web server the slower the server manager tasks become. We are a mix of OAS and WebLogic and I tested upgrading my test OAS machine with JRockit JDK and it appeared to increase the performance but I haven't had the chance to apply this to the Production servers yet. We aren't using any kind of load balancing either.

My ticket is no longer opened. I had it waiting on my completion of the JRockit JDK upgrade and during my vacation they auto-closed it. I tried to reopen it but they haven't done that yet. The support person tried to blame VMware but it worked fine before the tools release upgrade.

Sorry to dig up a relatively old forum topic but have you found a resolution to your Server Manager performance issue?

We still haven't and SM is becoming unusable (for the Web server User Sessions, caches etc). Our issue seems to be more related to the OAS servers. We are in the midst of moving to WebLogic but I can't see us getting the switch over completed anytime soon.