admin console refresh problem in websphere 5.1.1.1 - Websphere

This is a discussion on admin console refresh problem in websphere 5.1.1.1 - Websphere ; Here is my Environment Details- DeploymentManager: v5.1.1.1 ApplicationServer: v5.1.1.1 OS: RedHat8 When I click to the left pane of the console and select Application Servers or Enterprise Applications, to see the status on the right pane take exceedingly large amount ...

+ Reply to Thread
Results 1 to 7 of 7

Thread: admin console refresh problem in websphere 5.1.1.1

  1. admin console refresh problem in websphere 5.1.1.1

    Here is my Environment Details-

    DeploymentManager: v5.1.1.1
    ApplicationServer: v5.1.1.1
    OS: RedHat8

    When I click to the left pane of the console and select Application
    Servers or Enterprise Applications, to see the status on the right pane
    take exceedingly large amount of time, most of time more than 60 secs.

    Also when I make some changes and save via admin console, it takes more
    time than the normal.

    Here is what I have checked-

    1. Communication between the DM and AppServer box is fine and there is
    no firewall in between.

    2. The network cards are set to 100MB Full Duplex on both DM and App
    Server boxes.

    3. CPU/Memory utilisation is not a problem.

    Now I am wondering if anyone has experienced similar behavior in past or
    have some more ideas where I can look for to debug this slowness.

    Thanks in advance-

    Kailash

  2. Re: admin console refresh problem in websphere 5.1.1.1

    Hello,

    I've seen this kind of behavior when a Node Agent fails to respond in a
    timely fashion.

    When the list of Application Servers or Enterprise Applications builds up,
    can you correlate the slow responding ones to a particular Node ?

    Typically, I've seen the slow ones are located on the same machine and this
    indicates which machine may be in trouble. Alternatively, I also look at
    System Administration > Nodes to see which Nodes does not respond swiftly.

    Remains then to identify why this Node is not so responsive...

    HTH.



  3. Re: admin console refresh problem in websphere 5.1.1.1

    Ben_ wrote:
    > Hello,
    >
    > I've seen this kind of behavior when a Node Agent fails to respond in a
    > timely fashion.
    >
    > When the list of Application Servers or Enterprise Applications builds up,
    > can you correlate the slow responding ones to a particular Node ?
    >
    > Typically, I've seen the slow ones are located on the same machine and this
    > indicates which machine may be in trouble. Alternatively, I also look at
    > System Administration > Nodes to see which Nodes does not respond swiftly.
    >
    > Remains then to identify why this Node is not so responsive...
    >
    > HTH.
    >
    >


    Thank you for your suggestion.

    1. I have one Deployment Manager and one Application Server on separate
    physical machine.

    2. When the list of Application Servers or Enterprise Applications
    builds up there isn't any particular app thats slow, all takes almost
    same amount of time.

    3. Did you find the cause in your case?

    After reading through your post, I stopped the nodeagent on the
    appserver and tried to restart it, DM does not seem to recognise it for
    a while. I had to stop and start DM and nodeagent several times before
    they finally started seeing/talking to each other. This is weird. I do
    not see anything obvious in the deployment manager and nodeagent logs
    either.

    Thanks,
    Kailash

  4. Re: admin console refresh problem in websphere 5.1.1.1

    Hello,

    I use to see this when, on Windows, a process runs out of memory and the JVM
    generates a heapdump (automatic on Windows). This freezes the JVM and thus
    the Application Server. So the Node Agent waits for the Application Server
    to give its status and it slows down the Administrative Console.

    But maybe your problem is a bit different. Can you have a look at PQ92409
    and see if it helps ? (Without it, it takes ten minutes to stop and start an
    Application Server). Maybe you can simply apply the most recent Cumulative
    Fix (WAS 5.1.1.3, IIRC). There may be other fixes for issues with the
    Administrative Console.



  5. Re: admin console refresh problem in websphere 5.1.1.1

    Ben_ wrote:
    > Hello,
    >
    > I use to see this when, on Windows, a process runs out of memory and the JVM
    > generates a heapdump (automatic on Windows). This freezes the JVM and thus
    > the Application Server. So the Node Agent waits for the Application Server
    > to give its status and it slows down the Administrative Console.
    >
    > But maybe your problem is a bit different. Can you have a look at PQ92409
    > and see if it helps ? (Without it, it takes ten minutes to stop and start an
    > Application Server). Maybe you can simply apply the most recent Cumulative
    > Fix (WAS 5.1.1.3, IIRC). There may be other fixes for issues with the
    > Administrative Console.
    >
    >


    I have v5.1.1.1 working on separate environment without any problem.
    Anyway I am going to apply cumulative fix 3 and see if that helps.

    THanks,
    Kailash

  6. Re: admin console refresh problem in websphere 5.1.1.1

    Hi Kailash,
    I noted the same behavior with the security enabled.
    At this time I disabled the security and the admin console response time
    improved considerably.

    "Kailash Kayastha" wrote in message
    news:d21ie1$301e$1@news.boulder.ibm.com...
    > Here is my Environment Details-
    >
    > DeploymentManager: v5.1.1.1
    > ApplicationServer: v5.1.1.1
    > OS: RedHat8
    >
    > When I click to the left pane of the console and select Application
    > Servers or Enterprise Applications, to see the status on the right pane
    > take exceedingly large amount of time, most of time more than 60 secs.
    >
    > Also when I make some changes and save via admin console, it takes more
    > time than the normal.
    >
    > Here is what I have checked-
    >
    > 1. Communication between the DM and AppServer box is fine and there is
    > no firewall in between.
    >
    > 2. The network cards are set to 100MB Full Duplex on both DM and App
    > Server boxes.
    >
    > 3. CPU/Memory utilisation is not a problem.
    >
    > Now I am wondering if anyone has experienced similar behavior in past or
    > have some more ideas where I can look for to debug this slowness.
    >
    > Thanks in advance-
    >
    > Kailash




  7. Re: admin console refresh problem in websphere 5.1.1.1

    Ben_ wrote:
    > Hello,
    >
    > I use to see this when, on Windows, a process runs out of memory and the JVM
    > generates a heapdump (automatic on Windows). This freezes the JVM and thus
    > the Application Server. So the Node Agent waits for the Application Server
    > to give its status and it slows down the Administrative Console.
    >
    > But maybe your problem is a bit different. Can you have a look at PQ92409
    > and see if it helps ? (Without it, it takes ten minutes to stop and start an
    > Application Server). Maybe you can simply apply the most recent Cumulative
    > Fix (WAS 5.1.1.3, IIRC). There may be other fixes for issues with the
    > Administrative Console.
    >
    >

    Installing cumulative fix did not help. I forgot to mention earlier that
    I used the migration tools that comes with was v5.1 to migrate the
    application from v5.0.2 on this server. So I am guessing that something
    did not go right during the migration process. This is a development box
    and moreover, we need to change the OS from redhat to sles, so we are
    rebuilding the box. Anyway thanks for your tips.

    -Kailash

+ Reply to Thread