ACE slowness issue when one server goes down

Hi,
We are having two application servers.Both are load balanced using ACE.
When we bring down one server, we find that when we upload some files into the second application server, its too slow.
But when primary server comes up again the performance increases.This issue happens only when we bring the primary server down.
We are using cookie based stickiness.Any ideas where we can look into.
Rgds.,
Sachin

Depending on the load-balancing algorithm or predictor that you configure, the ACE performs a series of checks and calculations to determine which server can best service each client request. The ACE bases server selection on several factors including the source or destination address, cookies, URLs, HTTP headers, or the server with the fewest connections with respect to load.
http://www.cisco.com/en/US/docs/app_ntwk_services/data_center_app_services/ace_appliances/vA1_7_/configuration/slb/guide/classlb.html

Similar Messages

  • Automating the process of changing SAP servers when one server is down

    Hi,
    We have observed that sometimes users are unable to run a CRM IC web application due to server being down.
    So we want to automate the process that if a server is down then it will change to another server so that the application can run and the downtime is reduced.
    Can you please let me know if anyone has done anything on this front.
    Kindly send your suggestions.
    Thanks in advance.
    Regards,
    Jaisish Mohanty

    You can use sapwebdispatcher together with logon groups for this purpose:
    http://help.sap.com/saphelp_nw70ehp1/helpdata/en/be/6041418e998739e10000000a1550b0/frameset.htm
    Markus

  • Call drop when VXML server goes down

    Can somebody tell me what should ideally happen in the below given scenario?
    We have two media/vxml servers. We have a ACE load balancer which sends calls to these two servers. When the customer call is sent to CVP for self service (VXML application) , if the VXML server is shut down (via the CVP OAMP) , should the call drop ? or survivability kicks in the gateway ?
    What we saw is the call is dropped after 2-3 seconds of silence.
    We see in the CVP PIM logs that " called party disconnect" event and ends the call. Similarly in the gateway we see the transport down message and discnnects the call. is this correct?
    Is there any way to preserve this call and treat it?

    Hi Shravan,
    If you had configured survivability then ideally that call should hit the survivabilty. Normally on the VXML Gateways survivability is not auto configured, we would need to configure accordingly depending on the requirement. If survivability is not configured then the call would disconnect
    Additionally, could you please let us know the call flow for the self service in your setup
    Thanks,
    Dass
    Please rate useful posts

  • JMS Uniform Distribute Queue Unit Of Order, problem when one node goes down

    Hi ,
    I have the following code which post a message (with Unit of Order set ) to a Uniform Distribute Queue in a cluster with two member servers (server1 and server2).
    --UDQ is targeted to a subdeployment that is mapped to two JMS servers pointing to each member servers
    --Connection Factory is using default targeting ( i tried mapping to Sub deployment also)
    javax.naming.InitialContext serverContext = new javax.naming.InitialContext();
    javax.jms.QueueConnectionFactory qConnFactory = (javax.jms.QueueConnectionFactory)serverContext.lookup(jmsQConnFactoryName);
    javax.jms.QueueConnection qConn = (javax.jms.QueueConnection)qConnFactory.createConnection();
    javax.jms.QueueSession qSession = qConn.createQueueSession(false, Session.AUTO_ACKNOWLEDGE);
    javax.jms.Queue q = ( javax.jms.Queue)serverContext.lookup(jmsQName);
    weblogic.jms.extensions.WLMessageProducer qSender = (weblogic.jms.extensions.WLMessageProducer) qSession.createProducer(q);
    qSender.setUnitOfOrder("MyUnitOfOrder");
    javax.jms.ObjectMessage message = qSession.createObjectMessage();
    HashMap<String, Object> map = new HashMap<String, Object>();
    map.put("something", "SomeObject");
    message.setObject(map);
    qSender.send(message);
    } catch (Exception e) {           
    Steps followed:
    1. Post a message from "server1"
    2. Message picked up by "server2"
    3. Everything fine
    4. Shutdown "server2"
    5. Post a message from "server1"
    6. ERROR: "hashed member of MyAppJMSModule!MyDistributedQ is MyAppJMSModule!MyJMSServer-2@MyDistributedQ which is not available"
    WebLogic version : 10.3.5
    Is there a way (other than configuring Path Service ) to make this code work "with unit of order" for a UDQ even if some member servers go down ?
    Thanks very much for your time.

    If you want to avoid use of the Path Service, then the alternative is to make the destination members highly available. This will help ensure that the host member for a particular UOO is up.
    One approach to HA is to configure "service migration". For more information see the Automatic Service Migration white-paper at
    http://www.oracle.com/technology/products/weblogic/pdf/weblogic-automatic-service-migration-whitepaper.pdf
    In addition, I recommend referencing Best Practices for JMS Beginners and Advanced Users
    http://docs.oracle.com/cd/E17904_01/web.1111/e13738/best_practice.htm#JMSAD455 to help with WL configuration in general.
    Hope this helps,
    Tom

  • How to show  an error Page in Weblogic when the server is down

    Hi ,
    I heard that every container has the facility to show an error page when the server is down and (when the request has arrived at that time ).
    Please tell me how to configure this in BEA weblogic ??
    Thanks .

    Hi,
    From my understanding weblogic dose not provide the thing which you are asking for, reason is given below.
    - An application is deployed only on managed server one (MS-1)
    - Now if the request comes to the server (MS-1) when it was running and then for some reason MS-1 goes down then weblogic server would be able to send any response back to the client as its down.
    - Hence nothing can be done as the request object is been lost when the server goes down or even cant redirect to an error page as for doing that too the server has to be up and running.
    However there are other way round which can help you overcome this issue as shown below
    - Use Clusters with few managed servers in it which gives you high availability and load-balancing.
    - Use any web servers like Apache in front of weblogic servers.
    Using the above things you can also use customized HTTP Error Responses with error-page element in web.xml and create your own error pages which can be re-directed for the web server side also. For more information go through the below links
    Topic: Customizing HTTP Error Responses
    http://download.oracle.com/docs/cd/E13222_01/wls/docs92/webapp/configurejsp.html#wp157884
    Topic: error-page
    http://download.oracle.com/docs/cd/E13222_01/wls/docs81/webapp/web_xml.html#1017571
    Topic: ErrorPage
    http://download.oracle.com/docs/cd/E12840_01/wls/docs103/plugins/plugin_params.html#wp1155425
    Hope above information helps you.
    Regards,
    Ravish

  • Session Bean Client Hangs when one Server in Cluster Fails

    We are testing several failure scenarios and one has come up that concerns us.
    Some background: Were running a WLS6.1 cluster on two separate machines. We
    start a test client consisting of 50 active threads and let them start calling
    into a session bean. After a couple minutes we pull the network plug out of one
    of the machines to simulate an uncontrolled crash of the machine. Once the plug
    is pulled the clients hang and of more concern any new clients that we startup
    also hang. Has anyone successfully solved this problem?

    When we kill one of Weblogic instances in the cluster none of the clients fail.
    All of our clients fail-over to the remaining servers. It's pulling the network
    plug to our of the server that causes everything to hang. Not just or test client,
    but the other servers in the cluster hang. The control panel doesn't respond
    at all either. We currently have a support case open with BEA #348184 about this.
    We've gotten a prompt response in which we were asked to modify our configuration
    by deploying our beans to each individual server rather than the cluster. We
    did this, but the results so far have not changed.
    Thanks for the feedback,
    Howard.
    "Ade Barkah" <[email protected]> wrote:
    We haven't encountered something like that, so it could be a setup problem.
    Can you verify that the t3 url hostname the client threads use resolves
    to the
    ip addresses of each machine in the cluster? Are all machines in the
    cluster
    listening at the same port number? Also, does it matter if you kill one
    of the
    weblogic processes instead of pulling the plug? (i.e., if you leave the
    network
    layer up?)
    Check also that your threads aren't simply blocking each other when the
    server
    goes down? E.g. start multiple test client processes with one thread
    each just
    to test.
    What we notice is (with round-robin cluster policy), as we bring down
    one of
    the servers, the clients will continue to work on the second server,
    but will slow
    down between method invocations as they still attempt to connect to the
    downed
    server.
    After a short period of time (~30 seconds) the clients will fully switch
    to the
    second machine and processing continues at full speed again, until the
    downed
    machine is brought back up, at which point work is distributed evenly
    again.
    Also, when the first server is brought down, some of the clients may
    terminate
    with a PeerGoneException (or something similar to that.) So unless your
    threads
    are catching exceptions, they might terminate as well.
    regards,
    -Ade
    "howard spector" <[email protected]> wrote in message news:[email protected]...
    We are testing several failure scenarios and one has come up that concernsus.
    Some background: Were running a WLS6.1 cluster on two separate machines.We
    start a test client consisting of 50 active threads and let them startcalling
    into a session bean. After a couple minutes we pull the network plugout of one
    of the machines to simulate an uncontrolled crash of the machine. Once the plug
    is pulled the clients hang and of more concern any new clients thatwe startup
    also hang. Has anyone successfully solved this problem?

  • Server Goes Down Alert

    Hello,
    I am running SCOM 2012 R2 and I am trying to setup an alert to let me know when a server goes down.    I tried creating a rule that monitored the Management Server's "Operations Manager" Event Log for ID 20022.  I also configured
    the rule that parameter #2 must have my servers name.  I also tried EventDescription conatains my servers name.  None of them worked.  But I looked in the Event Log of my Management Server and as I looked at all of the 20022 entries, I noticed
    the computer I was on was listed as not heartbeating.  Two of my other servers that are currently on were listed as not heartbeating.  So obviously monitoring this 20022 entry is not reliable.  I do not want to setup a monitor for something
    like Ping Status.  I want to try to avoid using monitors because they could be taxing on the servers, my management server, and my network.

    Hi 
    yes event id 20022 is not reliable and never go for windows event id monitor for server shutdown or unavailability on agent , as scom health service will be stopped or connection closed with MS at the time of shutdown on agent.
    you need to go for ICMP port monitoring for server availability, as port monitoring run from remote server(watcher node) and can capture server unavailability and ping test.
    refer below link for port monitoring 
    http://technet.microsoft.com/en-us/library/hh457544.aspx
    http://www.systemcentercentral.com/custom-tcp-port-monitoring-management-pack-for-scom-2012/
    regards
    sridhar v

  • Failover did not happen when one node went down!!! PLEASE HELP

    Hi gurus,
    Yesterday one disaster struck my RAC database. We have two node cluster and it is 10.2.0.2, both of them located in different sites, yesterday suddenly power went down and the one of the network switch went down and got destructed, node one of RAC database was connected to that switch, but the failover did not happen to the node two as this should be the case when one node goes down the other should be available for all the node one sessions/connections.
    when I tried to ping/telnet the node 1, it was not happening because the switch was down, the network guyz connected the cables to other switch available. When I connected to the node 1, it was showing "Oracle is not available" message.
    And when I tried the other node, it was the same case but I did not see any error in alert log file. Then my TL restarted both the nodes and then the database was available.
    I am very confused that how the failover did not happen and how the database went down, PLEASE suggest something to how to identifiy what was happened. Thanks & Regards

    Thanks for your reply,
    after the network switch was replaced we connected to both the nodes and found that the instances are down with no reason given in the Alertlog file. We just restarted both the instances and then the database was up and the clients connected to both the instances with equal sessions on both the instances. I want to know that whether the failover can be done at the application side or it should be done on the database side i,e; in tnsnames.ora file with the required parameters? as in our scenario there is no failover configuration in the tnsnames.ora file.
    Thanks & Regards

  • Planning server goes down when I try to delete a member in Planning Web

    Hi,
    Planning server goes down when I try to delete a member in Planning Web. I am using Hyperion Planning 4 with Tomcat 4.1.30 and Windows Server 2003.
    I am not sure whats happening. This is the exception log I am looking at:
    2009-07-24 18:19:22 ApplicationDispatcher[HyperionPlanning] Servlet.service() for servlet jsp threw exception
    javax.servlet.ServletException
         at org.apache.jasper.servlet.JspServlet.service(JspServlet.java:249)
         at javax.servlet.http.HttpServlet.service(HttpServlet.java:853)
         at org.apache.catalina.core.ApplicationDispatcher.invoke(ApplicationDispatcher.java:684)
         at org.apache.catalina.core.ApplicationDispatcher.doForward(ApplicationDispatcher.java:432)
         at org.apache.catalina.core.ApplicationDispatcher.forward(ApplicationDispatcher.java:356)
         at org.apache.catalina.valves.ErrorDispatcherValve.custom(ErrorDispatcherValve.java:420)
         at org.apache.catalina.valves.ErrorDispatcherValve.status(ErrorDispatcherValve.java:327)
         at org.apache.catalina.valves.ErrorDispatcherValve.throwable(ErrorDispatcherValve.java:277)
         at org.apache.catalina.valves.ErrorDispatcherValve.invoke(ErrorDispatcherValve.java:179)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:641)
         at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:163)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:641)
         at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:480)
         at org.apache.catalina.core.ContainerBase.invoke(ContainerBase.java:995)
         at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:174)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:643)
         at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:480)
         at org.apache.catalina.core.ContainerBase.invoke(ContainerBase.java:995)
         at org.apache.coyote.tomcat4.CoyoteAdapter.service(CoyoteAdapter.java:199)
         at org.apache.coyote.http11.Http11Processor.process(Http11Processor.java:828)
         at org.apache.coyote.http11.Http11Protocol$Http11ConnectionHandler.processConnection(Http11Protocol.java:700)
         at org.apache.tomcat.util.net.TcpWorkerThread.runIt(PoolTcpEndpoint.java:584)
         at org.apache.tomcat.util.threads.ThreadPool$ControlRunnable.run(ThreadPool.java:683)
         at java.lang.Thread.run(Thread.java:534)
    ----- Root Cause -----
    java.lang.OutOfMemoryError
    2009-07-24 18:19:22 ErrorDispatcherValve[localhost]: Exception Processing ErrorPage[errorCode=500, location=/Error.jsp]
    javax.servlet.ServletException
         at org.apache.jasper.servlet.JspServlet.service(JspServlet.java:249)
         at javax.servlet.http.HttpServlet.service(HttpServlet.java:853)
         at org.apache.catalina.core.ApplicationDispatcher.invoke(ApplicationDispatcher.java:684)
         at org.apache.catalina.core.ApplicationDispatcher.doForward(ApplicationDispatcher.java:432)
         at org.apache.catalina.core.ApplicationDispatcher.forward(ApplicationDispatcher.java:356)
         at org.apache.catalina.valves.ErrorDispatcherValve.custom(ErrorDispatcherValve.java:420)
         at org.apache.catalina.valves.ErrorDispatcherValve.status(ErrorDispatcherValve.java:327)
         at org.apache.catalina.valves.ErrorDispatcherValve.throwable(ErrorDispatcherValve.java:277)
         at org.apache.catalina.valves.ErrorDispatcherValve.invoke(ErrorDispatcherValve.java:179)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:641)
         at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:163)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:641)
         at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:480)
         at org.apache.catalina.core.ContainerBase.invoke(ContainerBase.java:995)
         at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:174)
         at org.apache.catalina.core.StandardPipeline$StandardPipelineValveContext.invokeNext(StandardPipeline.java:643)
         at org.apache.catalina.core.StandardPipeline.invoke(StandardPipeline.java:480)
         at org.apache.catalina.core.ContainerBase.invoke(ContainerBase.java:995)
         at org.apache.coyote.tomcat4.CoyoteAdapter.service(CoyoteAdapter.java:199)
         at org.apache.coyote.http11.Http11Processor.process(Http11Processor.java:828)
         at org.apache.coyote.http11.Http11Protocol$Http11ConnectionHandler.processConnection(Http11Protocol.java:700)
         at org.apache.tomcat.util.net.TcpWorkerThread.runIt(PoolTcpEndpoint.java:584)
         at org.apache.tomcat.util.threads.ThreadPool$ControlRunnable.run(ThreadPool.java:683)
         at java.lang.Thread.run(Thread.java:534)
    ----- Root Cause -----
    java.lang.OutOfMemoryError
    Please do help me........ how to resolve this.
    Thanks.

    Hi,
    You could try increasing the java heap size, though sometimes an issue can occur if you are trying to delete a member when a workflow process has started, do you have one running?
    Cheers
    John
    http://john-goodwin.blogspot.com/

  • JMS Exception when server goes down (Urgent) ?

    Hi All,
    I have three topics in the JMS Server which is on the top of Web logic server 8.1.
    On the client side I have Swing thick cleint.For each topic I have a listener.
    When the web logic server goes down ( I shut down manually ) Only one listener is recieving the Exception and the Swing client reacts accordingly.
    After this ,
    Client continousoly checks whether the server is up or not. Once the server is up again the Exception is thrown that server is down ( I guess now the second listener is recieving the message).
    After some time again the third listener recives the message.....
    and the client reacts on that exception 3 times.
    How can I get rid of this ?
    thanks in advance.
    san

    Hi,
    this is probably a very good question to ask support. 'Logical' deployment sequence
    was added to the WebLogic only recently (6.1sp?) ('logical' means that first datasources
    are deployed, then EJB's and then web-apps, deployment-order parameter was added etc).
    So, for applications where shutdown sequence is important there must be a document which
    describes how server shutdown works.
    On the other hand, this may not be as important as deployment sequence, because running
    instance can crash at any time and it should not affect anything.
    Jen <[email protected]> wrote:
    Repost
    Hi,
    Most of my ejbs use connection from the connection pool. At the weblogic server shut
    down time (heavy load), I can see the jsps call the ejb, sometime jsp can't find
    the bean because it's removed from jndi. sometime jsp found the bean but when the
    bean try to talk to database, ejb can't use the connection because the datasource
    is removed from jndi. Also my shutdown class waits for the threads finish up, and
    the threads calls the ejb and do the db operation, and the conn is cut when the db
    operation is half way. So, the order of events when server shut down is important
    to me. I am wondering how other people deal with it. Thanks --
    Dimitri

  • Shouldn't ACE 4710 ignore cookie stickiness when the server is down?

    Hello,
    I have implemented sticky load balancing with cookies. The problem is that if one of my two servers in the server farm is down (and even if the ace recognizes it as down via a probe) it keeps sending the requests to the server that is down, obviously because it has set a cookie for this server,
    Shouldn't the ACE ignore the cookie when the server is down?
    Is there a command to ignore cookie stickiness if the server is down? Is there another workaround?
    an example of my config is
    serverfarm host SF_Ebanking
      rserver RS_IAS_1 XXXX
        conn-limit max 4000000 min 4000000
        probe http_probe_ebanking
        inservice
      rserver RS_IAS_2 XXXX
        conn-limit max 4000000 min 4000000
        probe http_probe_ebanking
        inservice
    sticky http-cookie ACE_COOKIE ebanking_sticky
      cookie insert
      replicate sticky
      serverfarm SF_Ebanking
      16 static cookie-value "server01" rserver RS_IAS_1
      24 static cookie-value "server02" rserver RS_IAS_2
    thanks,
    george

    This is not as obvious as you seem to believe.
    ACE will not select a server that is down !!!! Even if the cookie points to that server.
    What might be happening is that the connection from the browser to the ACE has not been killed, so when client sends a new request it reuses the existing connection and ACE does allow an existing connection to be maintain with a dead server by default.
    Try the command 'failaction purge' under the serverfarm.
    This should kill the active connections with the dead server and allow a new connection to be open with the other server even if the cookie points to the dead one.
    Regards,
    Gilles.

  • How can a JMS Listener Notified get when a Topic Server Goes Down?

    I'm in the process of looking into pub/sub model for a centeralized data source where vertical applications subscribe to a topic to get updates on person demographic information.
              I was playing with the JMS topic examples and noticed that when I have a Durable Subscriber listener (java main) and the topic Server goes down. It never gets notified. When I restart the Topic server and send msgs to the topic, the Durable Listener doesn't get them.
              Once I restart the Durable Listener (java main), it receives all the messages that were pending.
              I implemented the ExceptionListener and it never gets a shutdown msg either. I shutdown the Topic server via Ctrl-C, Graceful Shutdown and Forced Shutdown.
              Is there a way for the client listener to determine when the Topic Server went down, came backup, and restore the connection to it automatically? Or do we have to implement some type of ping mechanism on the client listener and re-create the connection ourselves?
              Thanks. Later...
              - Wayne
              Edited by wlau1000 at 03/07/2007 8:34 AM

    Hi Wayne,
              I think the exception listener should have fired. Some comments/questions:
              - Does your app call "connection.setExceptionListener" in addition to implementing ExceptionListener?
              - Which version and maintenance/service pack are you using?
              - Is the client using the wl* "thin" client jars or the "thick" client jar?
              - Version 9.2 provides an automatic reconnection capability for consumers. This behavior isn't enabled by default due to natural side effects that some apps might not be prepared to handle (needles to say, I recommend reading the doc thoroughly).
              http://edocs.bea.com/wls/docs92/jms/recover.html#wp1320125
              Tom

  • Is there a  requirement to restart the JMS/Weblogc Server everytime when the DB goes down?

    Question:
    When the Database is shutdown gracefully everything was fine.
    But when the Database is brought down with shutdown abort the MDB consumer will never consume messages from the topic again. And we need to restart the WebLogic Server in order to get the messages consumed again.
    So is there a requirement to restart the JMS/Weblogc Server everytime when the DB goes down? If not, what type of failures will require the JMS/WLS restart?

    On a DB failure, a WL JDBC store service will make a brief attempt to reconnect before shutting itself down along with any services that depend on the store. 
    It isn't necessary to restart the entire WebLogic Server JVM to bring the affected service(s) back if you can use the Automatic Service Migration feature.  ASM can automatically restart a failed service on a different WL Server in the same cluster, and/or can try a restart-in-place for the service if the service's original host WL Server JVM is still running.   The Automatic Service Migration (pdf) white-paper has a thorough discussion of this area.
    In addition to ASM, there's also a "whole server migration" option that can automatically restart or migrate an entire WL server.
    Tom

  • Redirect to a custom error page when Forms Server shuts down

    Hi,
    I want the forms application to redirect to a custom error page whenever Forms Server goes down. (like network issues, unscheduled shutdown...) Whenever this happens, the user gets a pop up with FRM messages and user has to acknowledge the pop up. Since the application is used in the shop floor, there will not be any user always in front of the screen and will be a very big issue as there are at least 70-80 PC's on the floor which will be attended fewer times(only when there is issue with the machine) other wise the application displays important events of the machine,etc.,
    I am trying to develop a simple HTML error page(may be with applet) which will keep querying the forms server and once the forms server is up and running, the error page will revert back to the application which the user was working earlier to shutdown. Other option is the redirect to main menu of the application (this may be a PSP or a JSP or a Oracle Forms Menu ).
    Is there any event in forms to capture the loss of connection to forms server ??
    BTW, I am using Oracle Application Server 10g Forms and Reports on Windows 2000 Server...
    Any ideas...
    Thanks in advance
    Purush

    I don't think it will be easily possible in Forms for a couple of reasons.
    1. In Forms you do not have query rights, and as soon as you want to edit data, you have to log in. In Forms you log in once, and this sets your rights for the whole session. You will have to do some programming in Forms to achieve what you want. E.g. as soon as you hit the Save button you are asked for a username and password.
    2. A refresh cannot be done in Forms. Remember that Forms is a Java applet. The URL in the browser is just pointing to the startup page only. If you refresh your screen, you are taken back to the login screen of Forms again.
    Well, there will probably be ways around these problems. I'm thinking in the line of calling every form in a separate URL instead of starting a Forms application where one form is called from another form. The default 'query only' form can be called from a startup html page. This page can be refreshed. As soon as the user wants to update data, he clicks on a link or button and is taken to a login form.
    If the refresh is unsuccessful you are taken to a custom error page, like in your ASP application.
    A problem with an automatich refresh may be hanging sessions. Forms server does not always kill sessions.
    And also, how would you avoid a refresh if a user is currently using the form?

  • Sticky issue when failed server become active

    Hi,
    We are operating the ace(csm) and using the sticky option.
    Assuming that We have two servers with sicky for redundancy and being opreated active - active, 
    I realized that if one server go down and go up, Csm dosen't forword packet to restored server.
    I think it is due to the sticky option. We need csm to play loadbalancing two servers once one is restored immediately.
    Is there anything opinion or solution?
    Thanks,

    If a real goes down (fails probe) then the csm will remove it from service and not send connections to it. The entry will also be removed from the sticky table. When it comes back, it will receive new connections when it is elgible. If all incoming connections match existing sticky entries then it will take time to gather load. If you configure "predictor leastcons" under the serverfarm and limit the sticky life, then it will allow the failed real to recover load quicker. Matthew

Maybe you are looking for