Caching in IFS

As an application developer I would like
to understand what all iFS caches. What
happens in a load balanced scenario?
Let us say I have two iFS applications
that hit the same iFS database. If iFS
does any caching, does it take care of
refreshing the cache when one instance
of the application changes data (user
profile, ACLs etc).
I would like to know what gets cached
in iFS.
null

Allow me to clarify a few things.
You can think of an iFS service as the "black box" that handles all the communication between the iFS server software (Java) and the database. An iFS service is "pointed to" a specific schema in a specific database instance (recall that you get to choose these with the iFS Configuration Assistant).
The iFS service manages the caches, connection pools, etc., hiding all the details from iFS clients. (iFS clients are you, the developer who uses the published API, as well as each protocol server and application, such as FTP and the WebUI.)
In iFS 1.0 and 1.1, each iFS application (like ServerManager or SMB) has a single iFS service -- that is, each JVM has a single iFS service (note that this is not a restriction, it's just the default configuration; if you really wanted to, you could have multiple iFS services in a single JVM).
Each iFS service has an associated configuration file, which you can find in the $ORACLE_HOME/ifs/settings/oracle/ifs/server/properties directory. Note that just because a service configuration file is present in the afforementioned directory does not mean that that service is up and running; the service properties file is just one of the services that you COULD start up.
Each iFS service manages a set of iFS sessions (you create a session when you get a new LibrarySession instance using the LibraryService.connect method).
When you start adding middle tiers, the pattern still holds. For instance, if you have FTPServer1 on MachineA, that's one iFS service. If you add FTPServer2 on the same machine (MachineA), that's another iFS service. If you add FTPServer3 on MachineB (a second computer), that's just another service. The visibility rules described above apply in the same way in all of these cases. All you need is a connection to the database, which is provided by the Oracle Client (JDBC over Net8).
Finally, let me just say that synchronizing these caches is non trivial -- iFS is an extremely powerful distributed application. (I had to get that in there.)

Similar Messages

  • Does anyone have a Portal working with 50 or more Users concurrently ?

    I WOULD LIKE TO KNOW IF ANYBODY HAS A PRODUCTION PORTAL WITH 50 OR MORE USERS WORKING CONCURRENTLY.
    I AM INTERESTED TO KNOW HOW PORTAL PERFORMS WITH MANY USERS.
    ANY ADVICE OR SUGGESTIONS WILL BE VERY WELLCOME.
    TKS!

    Without knowing your hardware setup, your functional requirements, and the user base, and can't answer that question with 100% certainty. But, I can tell you that a multiple server configuration will definitely provide you with more performance and scalability (assuming proper configuration).
    As far as using Standard Edition vs Enterprise Edition, there will be no differences from a performance/scalability perspective. With Enterprise edition you get some tools you don't get with Standard Edition such as WebCache, Forms, Reports, Discoverer, 8i Cache (database cache), and iFS. I'm not certain if you get OID with both editions, but I think you might. If you dont have OID you wont be able to use the LDAP authentication features of Portal (unless you already have your own LDAP directory service set up).
    The product lists are from my memory (not always dependable), so you should verify those through your sales representative before making any decisions.
    Hope this helps...
    null

  • Surrogate-Control http header

    Access to files from the http viewer seems to include some Surrogate-Control directives in the http response header. These directives either give a max-age of 0 or specify no-store.
    Is this configurable for iFS and if so how can I avoid this. To be able to cache read-only documents would speed up access tremendously.

    You may be running into an SSO bug where SSO is inserting these "no-cache" headers. (iFS does not insert any extra headers on its own.) Try posting on the SSO message board or following up with Oracle support to get the latest patches for SSO.

  • Dmenu_run color scheme

    Hey guys i have been searching for ways to change the color scheme in dmenu_run (/usr/bin/dmenu_run) and found the solution.
    here is my dmenu_run file:
    #!/bin/sh
    CACHE=${XDG_CACHE_HOME:-"$HOME/.cache"}/dmenu_run
    IFS=:
    if test "`ls -dt $PATH "$CACHE" 2> /dev/null | sed 1q`" != "$CACHE"; then
    mkdir -p "`dirname "$CACHE"`" && lsx $PATH | sort -u > "$CACHE"
    fi
    exe=`dmenu -b -fn '-*-helvetica-medium-r-normal-*-17-*-*-*-*-*-*-*' -nf '#e04613' -nb '#3b3b3b' -sf '#e04613' -sb '#000000' "$@"< "$CACHE"` && exec sh -c "$exe"
    This color scheme matches the Elegant brit theme.
    -nf is for color font
    -nb is for background
    -sf is for the font of highlighted text
    -sb is for the background of the highlight
    enjoy!
    Last edited by sudohalt (2011-11-28 00:15:12)

    hmmm... do you have an identity already? logo, etc? maybe
    some color in there could serve as the base?
    also, how about thinking about this from a historical
    perspective.... every period has a "color history" – popular
    colors used during that era... what were the colors used in that
    time period in your region? have you visited some of the
    buildings/homes that are protected in that city's preservation
    society? what are the common colors in them? take photos and use
    them as base colors for kuler themes, creating triads and compund
    themse around them.

  • Caching problem with iFS web server / index.html

    There seems to be a problem with the iFS Java web server serving cached index.html documents after they have been modified.
    This only happens when the document is referenced by the folder name (e.g. /home/kennu/). If you reference it by the full name (/home/kennu/index.html), the new modified version comes up alright.
    I wonder if this is a bug or some kind of a feature? Is there an quick and easy workaround?
    Byt the way, is this the correct place to post bug reports? I didn't see iFS mentioned in the OTN Bug Database.
    null

    well, this is certainly not an intended effect. i'll look into it. in the meantime, the workaround already posted should pull you through.
    sorry about the inconvenience.
    -ilya
    <BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Kenneth Falck ([email protected]):
    There seems to be a problem with the iFS Java web server serving cached index.html documents after they have been modified.
    This only happens when the document is referenced by the folder name (e.g. /home/kennu/). If you reference it by the full name (/home/kennu/index.html), the new modified version comes up alright.
    I wonder if this is a bug or some kind of a feature? Is there an quick and easy workaround?
    Byt the way, is this the correct place to post bug reports? I didn't see iFS mentioned in the OTN Bug Database.<HR></BLOCKQUOTE>
    null

  • IFS and Data Cache Configuration

    Hi!!
    i installed 9iAs Enterprice Edition on my machine that already had Oracle db 8.1.7, i'm trying to set up ifs, but when i choose the option "Oracle8i is on THIS machine" in the ifs wizard i've got a protocol adapter error, if i choose the option "oracle8i on another machine" i solved the problem, why is this happening???
    the other problem it's with theb database cache i started the wizard, but in 16% when it's creating the cache it stoped it didn't show any error, i leaved for 2 days but it just reached the 16% and stayed ther.
    could you please help me..

    For iFS configuration on Enterprise Edition you can't select the option, "Oracle8i is on THIS machine". Actually this is a typo. This should be "Oracle8i is on THIS oracle_home". This will be fixed in future releases. iFS can't be configured against database cache(icache) also. So definitly you need a database which is on another oracle_home and you should select the second option, "oracle8i on another machine".
    From the second problem, I assume you are installing 9iAS on a Windows NT platform. On NT, sometimes database cache hangs. I assume it is mainly because of less memory. How much is the memory in your system? Are you running so many processes on the system while cache configuration? If so, reduce it as much as you can and retry for cache configuration. You may need to manually shutdown one SQLPLUS.EXE process in 'Task manager' started from the current cache configuration before retrying the config.

  • Please help: ORA-04030: Out of Process memory error when starting 9.0.1 iFS

    When starting the 9.0.1 iFS running on a 9.0.1.4 Database and Tru64 5.1 O/S, the following errors are displayed:
    IFS-35205: Failed to get prepared statement
    java.sql.SQLException:ORA-04030: Out of process memory when trying to allocate 1048 bytes (callheap,kghsstk).
    Database has the following parameters:
    shared pool 150M
    java pool 100M
    large pool 5M
    DB Cache 60M
    DB Block Size 8K
    Sort area size 1M
    Server has 1GB ram, 12GB swap.
    Please can anyone help?

    Hi Alan,
    your help with this is most appreciated!
    I ran your suggested generated SQL script, but at first the same problem was occuring with the first ALTER statement that happened to be generated.
    However, your explanation mentioned the view 'odm_publicobject' and although this view wasn't in the resulting SQL, I found that compiling the similarly named views in this order worked...
    ODMV_PUBLICOBJECT
    ODBV_PUBLICOBJECT
    ODMBS_PUBLICOBJECT
    ODMS_PUBLICOBJECT
    ODUS_PUBLICOBJECT
    ODUV_PUBLICOBJECT
    (Perhaps other combinations would have worked, but in anycase certainly some orders didn't work...)
    At this point, running the complete original generated SQL script then worked successfully and low and behold as you predicted, IFS started successfully.
    (I didn't actually bother removing these 6 already re-compiled views from the original script, so they got re-compiled twice...)
    So this means that we can now actually stop and start IFS as required without re-installing IFS everytime - which is obviously great!
    I will certainly be watching out for 9.0.3, but my understanding for the reason we were going with 9.0.1 in the first place was that this was included in the RDBMS CD-pack, whereas I believe from 9.0.2 onwards it was seperately licensed. I will check with our Oracle dba's.
    But at least we can prove in principle now whether it will suit our requirements....
    Once again, many thanks for looking into this at such a deep level - very much appreciated.
    Graham.

  • Ifs and rmi

    hi,
    i wrote a rmi-server which accesses the ifs-repositry and retrievs the folderlist (similar to the "jdb error" post). now i can call the server-method "RmiServer.list(...)" which returns a string containing some folder-names. but: trying to call that RmiServer.list() one more time causes the client to hang. i took a look at the servers thread dump and notices one differenc in the thread-lists before and after the ifs-access.
    one thread seems to loop/deadlock/hang/../?.
    here is the thread dump after the ifs-call:
    Full thread dump:
    "LeaseChecker" (TID:0xec99eb90, sys_thread_t:0x154888, state:CW, thread_t: t@12, sp:0xef080ea0 threadID:0xef081dd8, stack_base:0xef081d6c, stack_size:0x20000) prio=5
    sun.rmi.transport.DGCImpl$LeaseChecker.run(DGCImpl.java:303)
    java.lang.Thread.run(Thread.java)
    "KeepAlive" (TID:0xec99d528, sys_thread_t:0x108020, state:CW, thread_t: t@9, sp:0xef110ea0 threadID:0xef111dd8, stack_base:0xef111d6c, stack_size:0x20000) prio=5
    sun.rmi.transport.KeepAlive.run(ObjectTable.java:182)
    java.lang.Thread.run(Thread.java)
    "Reaper" (TID:0xec99d568, sys_thread_t:0xe2d28, state:CW, thread_t: t@7, sp:0xef140e40 threadID:0xef141dd8, stack_base:0xef141d6c, stack_size:0x20000) prio=5
    sun.rmi.transport.Reaper.run(ObjectTable.java:199)
    java.lang.Thread.run(Thread.java)
    "TCP Accept-1" (TID:0xec99d198, sys_thread_t:0xf5e10, state:CW, thread_t: t@6, sp:0xef170ea0 threadID:0xef171dd8, stack_base:0xef171d6c, stack_size:0x20000) prio=5
    sun.rmi.transport.tcp.TCPTransport.retryServerSocket(TCPTransport.java:344)
    sun.rmi.transport.tcp.TCPTransport.run(TCPTransport.java:430)
    java.lang.Thread.run(Thread.java)
    "SIGQUIT handler" (TID:0xec9982a0, sys_thread_t:0x8b3b8, state:R, thread_t: t@5, sp:0xef1f1af0 threadID:0xef1f1dd8, stack_base:0xef1f1d6c, stack_size:0x20000) prio=0 current thread
    "Finalizer thread" (TID:0xec998088, sys_thread_t:0x8b328, state:CW, thread_t: t@4, sp:0xef341a30 threadID:0xef341dd8, stack_base:0xef341d6c, stack_size:0x20000) prio=1
    "main" (TID:0xec9980b0, sys_thread_t:0x724f8, state:CW, thread_t: t@1, sp:0xefffdad8 threadID:0x24190, stack_base:0x0, stack_size:0x800000) prio=5
    Monitor Cache Dump:
    <unknown key> (0x154888): <unowned>
    Waiting to be notified:
    "LeaseChecker" (0x154888)
    <unknown key> (0xf5e10): <unowned>
    Waiting to be notified:
    "TCP Accept-1" (0xf5e10)
    sun.rmi.transport.tcp.TCPTransport@EC99CEF8/ECA883D8: owner "TCP Accept-1" (0xf5e10, 1 entry)
    <unknown key> (0xe2d28): <unowned>
    Waiting to be notified:
    "Reaper" (0xe2d28)
    <unknown key> (0x108020): <unowned>
    Waiting to be notified:
    "KeepAlive" (0x108020)
    Registered Monitor Dump:
    Thread queue lock: owner "SIGQUIT handler" (0x8b3b8, 1 entry)
    Waiting to be notified:
    "main" (0x724f8)
    Name and type hash table lock: <unowned>
    String intern lock: <unowned>
    JNI pinning lock: <unowned>
    JNI global reference lock: <unowned>
    BinClass lock: <unowned>
    Class loading lock: <unowned>
    Java stack lock: <unowned>
    Code rewrite lock: <unowned>
    Heap lock: <unowned>
    Has finalization queue lock: <unowned>
    Finalize me queue lock: <unowned>
    Waiting to be notified:
    "Finalizer thread" (0x8b328)
    Monitor registry: owner "SIGQUIT handler" (0x8b3b8, 1 entry)
    comparing the thread dumps before and after the ifs-access differs in one thread: the rmi thread "TCP Accept-1". above shown is the dump afterwards. before ifs-access the "TCP Accept-1" thread stack is as follows:
    "TCP Accept-1" (TID:0xec99d198, sys_thread_t:0xf5e10, state:R, thread_t: t@6, sp:0xef171040 threadID:0xef171dd8, stack_base:0xef171d6c, stack_size:0x20000) prio=5
    java.net.PlainSocketImpl.accept(PlainSocketImpl.java:379)
    java.net.ServerSocket.implAccept(ServerSocket.java:198)
    java.net.ServerSocket.accept(ServerSocket.java:181)
    sun.rmi.transport.proxy.HttpAwareServerSocket.accept(HttpAwareServerSocket.java:70)
    sun.rmi.transport.tcp.TCPTransport.run(TCPTranspo rt.java:376)
    java.lang.Thread.run(Thread.java)
    server-system: solaris 2.6, jdk: ifs' own jdk.
    client-system: nt-4.0, jdk: sun jdk 1.1.6
    i would be very happy for any help, so thanks in advice,
    alex

    I don't think we (Oracle) can invest the amount of time it would take to help you with this problem, since it involves RMI, which is not an Oracle product.
    My only (somewhat uninformed) thoughts are that there may be problems with the RMI itself.
    We have seen many problems with JDK 1.1.6, and recommend you using JDK 1.1.8.
    Perhaps that might help.

  • Memory Leak in iFS 1.1.9 (or JDBC) ?

    I suppose it's fair to ask why you'd want to reuse a library
    session. And, given what I've seen on Windows NT and 2000, I
    don't recommend it.
    It appears that reusing a library session leaks memory. I've
    not identified exactly where the leak is. I've only observed
    limitless memory consumption with frequent access to iFS and
    reusing a library session to retrieve objects or execute
    queries. Closing the session after each use seems to minimize
    or eliminate the apparent leaks.
    Is this a known problem?

    yes, this was a known problem in 1.1.9, and fixed in 9.0.1. The
    "leak" was in iFS (not JDBC), and caused by one of our caches
    being unbounded.
    sorry for the trouble-
    dave

  • IFS map drive

    Almost the same problem than cilent util()...
    The local 'o' drive has not been created...
    what's wrong?
    1. I am running utilities on NT machine,with window explorer.
    2. I can find hosting server in network neighborhood and "find computer" facility.
    3. The client machine is member of a NT domain
    4.I can see all machines in domain which contain iFS hosting server.
    What should I do?
    thank you very much...

    This is the spec of my machine:
    OS: Windows NT 4
    Memory: ~261 K RAM
    Processor: x86 Fam 6 Mod 8 Stepp 3 GenuineIntel ~648 Mhz
    SWAP: ~273 K
    NTFSServerlog file
    Wed Aug 23 16:50:07 CEST 2000
    iFS NT File Server Version 1.0.8.1.0 Production
    Wed Aug 23 16:50:08 CEST 2000
    no ifsproxy in shared library path
    Cache performance for session cache
    CurrentSessionCount=0
    HighestSessionCount=0
    CurrentSessionCache=0
    null

  • IfsException (IFS-10811) starting the example LoggingAgent

    Hi,
    I am trying to write my own Agent for iFS 9.0.1.
    The compilation goes fine but when trying to launch it in standalone mode,
    an exception occurs.
    To be sure it is not an implementation problem, I tried to launch the
    LoggingAgent provided in the "samplecode" folder. I get the same error:
    10/10/02 4:22 PM LoggingAgent: Requested to start
    10/10/02 4:22 PM LoggingAgent: Starting
    10/10/02 4:22 PM LoggingAgent: Start request
    10/10/02 4:22 PM LoggingAgent: Pre-run failed
    10/10/02 4:22 PM LoggingAgent: oracle.ifs.common.IfsException
    oracle.ifs.common.IfsException: IFS-45360: Unable to connect default session
    oracle.ifs.common.IfsException: IFS-45206: Unable to get credential
    oracle.ifs.common.IfsException: IFS-21008: Unable to connect to iFS service
    oracle.ifs.common.IfsException: IFS-10811: Unable to construct LibraryObject Cache for
    class oracle.ifs.server.cache.BoundedLibraryObjectCache (invalid class path specification)
    at oracle.ifs.server.S_LibrarySession.constructBoundedLibraryObjectCache(S_LibrarySession.java:3449)
    at oracle.ifs.server.S_LibrarySession.<init>(S_LibrarySession.java:1900)
    at java.lang.reflect.Constructor.newInstance(Native Method)
    at oracle.ifs.server.S_LibraryService.constructSession(S_LibraryService.java:2575)
    at oracle.ifs.server.S_LibraryService.connect(S_LibraryService.java:2368)
    at oracle.ifs.beans.LibraryService.connect(LibraryService.java:977)
    at oracle.ifs.management.domain.Server.getCredential(Server.java:374)
    at oracle.ifs.management.domain.IfsServer.connectSession(IfsServer.java:1034)
    at oracle.ifs.examples.servers.LoggingAgent.preRun(LoggingAgent.java:150)
    at oracle.ifs.management.domain.IfsServer$ServerRunner.run(IfsServer.java:2123)
    10/10/02 4:22 PM LoggingAgent: Stopping
    10/10/02 4:22 PM LoggingAgent: Timer stopped
    10/10/02 4:22 PM LoggingAgent: Stop request
    10/10/02 4:22 PM LoggingAgent: Timer stopped
    10/10/02 4:22 PM LoggingAgent: java.lang.NullPointerException
    java.lang.NullPointerException
    at oracle.ifs.examples.servers.LoggingAgent.disableEventListening(LoggingAgent.java:320)
    at oracle.ifs.examples.servers.LoggingAgent.postRun(LoggingAgent.java:198)
    at oracle.ifs.management.domain.IfsServer$ServerRunner.run(IfsServer.java:2155)
    10/10/02 4:22 PM LoggingAgent: Stopped
    The message complains about an "invalid class path specification". However,
    I am sure the class "oracle.ifs.server.cache.BoundedLibraryObjectCache" is
    present in the classpath (in the repos.jar library).
    Is this problem related to cache size ? If so, how can I increase the cache ?
    I know the Agent reads its configuration file (LoggingAgent.def) correctly and
    uses the appropriate service name and schema password because the main method
    succeeds to start the service and to initialize the LoggingAgent instance.
    The failure happens when the connectSession() method is invoked in the
    preRun() method.
    Any help will be greatly appreciated,
    Tristan

    I hate to suggest this, but this impliens that the IFS repos did not install properly. Can you try re-running the ifsconfig process and either create a new repos, or re-initialize your existing repos (Be careful re-initializing an existing repos will cause all info in the repos in the lost).

  • Error message IFS-12200, why it appears everytime?

    Hi
    What is the reason why the error message IFS-12200 keeps on appearing whenever I click the user name in the navigator of Oracle iFS Manager, even the user "system".
    I logged into the Oracle iFS Manger as amdin mode.
    Thanks a lot.
    null

    Your central cache file seems corrupted and you ave to delete this manual. Regarding your screenshot you seem to be on a Mac.
    Per OSX 10.7 the user library is hidden, if you don't have your user library in view use menu go in Finder with option key to reveal this library.
    Quit Bridge, go to your user library (user (your username) / library / Caches/ Adobe/ Bridge CSxx (your version) and inhere are a plug in file and a folder called Cache.
    Drag both files outside of the library (preferable to the trash) and restart Bridge. This creates both deleted files again but the cache folder is empty and you need to recache your files again (takes a bit of time) but then you should be OK.

  • IFS performance tuning

    I have oracle 9iFS setting on a windows box. It works well initially, after I batch loaded about 100,000 documents into it. It becomes quite slow. Open a folder through SMB client takes tens of seconds. Does anyone have some performance tuning tips? Or anyone has similiar situation with large amount of documents?
    Thanks

    I have oracle 9iFS setting on a windows box. It works well initially, after I batch loaded about 100,000 documents into it.It becomes quite slow. Open a folder through SMB client takes tens of seconds. Does anyone have some performance tuning tips?
    Or anyone has similiar situation with large amount of documents?
    Thanks If running analyze doesn't help the problem, please post a new thread, and we'll try to help you. When you do, please answer these questions:
    - How many documents per folder?
    - Can you detect whether your iFS Java processes or your Oracle processes are the bottleneck?
    + If the bottleneck is an 9iFS Java process, then go into the Enterprise Manager tool (configured with iFS), and bring up the Node Performance Dialog (see page 2-25 of the 9iFS Setup and Administration Guide).
    + For your server, bring up the information described in Figure 2-20 of the 9iFS Setup and Admin Guide. "Service Details: Committed Data Cache".
    - Adjust these settings higher and see if your performance improves.
    Alan

  • IFS Drive does not mount as well as Item Name error

    I'm having two problems:
    1. The IFS drive does not mount. (Log info is appended at the end)
    2. In IFS Manager, when I click on any of the users, I get the error: IFS-12200: Invalid item name specified (AnonymousCredentials)
    Fri Aug 04 11:44:26 CDT 2000
    iFS NT File Server Version 1.0.8.1.0 Production
    Fri Aug 04 11:44:26 CDT 2000
    ServerName(id): NtfsServer(2924)
    server: NtfsServer
    username: system
    AnonymousUsername: null
    DriveLetter: o
    NtfsSessionCache: 40
    OptimizeStartup: true
    LogLevel: 1
    LogStackTrace: true
    net start OracleIfsd
    Return code = 2
    ifsmount o
    Return code = 2
    SR Event: ID:3102 PENDING SPECIAL SUBTYPE:13 3106L
    SR Event: ID:3102 COMMIT SPECIAL SUBTYPE:13 3106L
    (0) Unable to open IFS device
    ifsdismount o
    Return code = 2
    Cache performance for session cache
    CurrentSessionCount=0
    HighestSessionCount=0
    CurrentSessionCache=0
    Thanks!
    null

    Fixed both problems:
    - Problem 1 was being caused because of insufficient memory
    - Problem 2 was fixed by re-installing IFS.
    null

  • IFS Node hanging (Needs Database restart)

    Hi ,
    We have system which was working pretty good.Recently we started getting these problems with the iFS node. We are not able to connect to iFS and our client applications are getting the following error.
    We have the follwing configuration.
    1) Oracle iFS 9.0.1
    2) Oracle DB 9.0.1
    3) We use iFS SDK as middle tier and run iFS service on middle tier
    4) We have some custom agents running on serverside (Node).
    5) Node runs with default service which uses "SmallServiceConfiguration".
    6) Clients also use service property file which has "SmallServiceConfiguration".
    Here is the exception which we are getting.
    java.lang.Exception: Exception while trying to get Connection
    oracle.ifs.common.IfsException: IFS-21008: Unable to connect to iFS service
    oracle.ifs.common.IfsException: IFS-10151: Unable to perform authentication
    oracle.ifs.common.IfsException: IFS-10621: Unable to acquire library connection
    oracle.ifs.common.IfsException: IFS-10630: No unallocated connections (timed out waiting)
         at oracle.ifs.server.ConnectionPool.acquireLibraryConnection(ConnectionPool.java:580)
         at oracle.ifs.server.IfsCredentialManager.authenticate(IfsCredentialManager.java:237)
         at oracle.ifs.server.DirectoryService.authenticate(DirectoryService.java:527)
         at oracle.ifs.server.S_LibraryService.connect(S_LibraryService.java:2495)
         at oracle.ifs.beans.LibraryService.connect(LibraryService.java:977)
         at com.cisco.as.ifs.util.CiscoService.connect(CiscoService.java:79)
    Looking forward for your reply.

    Hi Luis,
    We have web applications which creates library session for each user who logs in. The library session gets disconnected when users HttpSession gets invalidated.
    We did check how many connections were open from the application by running query against V$session table and there were around 30 inactive connections.
    Rightnow here are the details of properties which are set.
    # Service data cache
    IFS.SERVICE.DATACACHE.Size=7500
    IFS.SERVICE.DATACACHE.NormalTrigger=5000
    IFS.SERVICE.DATACACHE.UrgentTrigger=5500
    IFS.SERVICE.DATACACHE.EmergencyTrigger=6000
    IFS.SERVICE.DATACACHE.PurgeTarget=4000
    # Writeable connection pool
    IFS.SERVICE.CONNECTIONPOOL.WRITEABLE.MinimumSize=2
    IFS.SERVICE.CONNECTIONPOOL.WRITEABLE.TargetSize=5
    IFS.SERVICE.CONNECTIONPOOL.WRITEABLE.MaximumSize=10
    IFS.SERVICE.CONNECTIONPOOL.WRITEABLE.TargetSizeTimeout=1000
    IFS.SERVICE.CONNECTIONPOOL.WRITEABLE.MaximumSizeTimeout=10000
    # Readonly connection pool
    IFS.SERVICE.CONNECTIONPOOL.READONLY.MinimumSize=2
    IFS.SERVICE.CONNECTIONPOOL.READONLY.TargetSize=5
    IFS.SERVICE.CONNECTIONPOOL.READONLY.MaximumSize=10
    IFS.SERVICE.CONNECTIONPOOL.READONLY.TargetSizeTimeout=1000
    IFS.SERVICE.CONNECTIONPOOL.READONLY.MaximumSizeTimeout=10000
    These are values from "SmallServiceConfiguration" which we are using. Since we have web based application there might be time when some operations are executing concurrently. Do you think that has anything to do with Node being hung? .
    We haven't tried using "medium" / "Large" service configuration yet , but are thinking to change it to "medium" . Should we do the change at the middle layer where we run our webapplication or at Node layer ?
    Let me know your responses and we will try it and see whether that solves the problem.
    Thanks

Maybe you are looking for