RMI gcInterval - Does it invoke a Full GC?

My query is with regard to the bug reported here: [http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6200091]
The Bug description suggests that if the rmi gc interval is left at its default value of 1 minute (for JDK 1.5) it causes a Full Garbage Collection (and not a minor GC) cycle to trigger every minute.
Q: Is this understanding correct?
If the answer to the above is 'Yes', then,
Q: Does the full GC triggered whenever the 'sun.rmi.dgc.server.gcInterval'  elapses cause a full GC on the full JVM heap on the server and also have a FULL GC logged in the catalina.out (if running on Tomcat servlet container) similar to the following? I know the answer to this would be Yes! but I just want to be 200% sure!
[Full GC 36242.268: [Tenured: 284568K->284877K(1864192K), 1.9276716 secs] 307279K->284877K(2090304K), [Perm : 61047K->61047K(61184K)], 1.9280748 secs]Obviously assume that the following specifc JVM arguments in place:
-XX:+PrintGCTimeStamps
-XX:+PrintGCApplicationConcurrentTime
-XX:+PrintGCApplicationStoppedTime
-XX:+PrintGCDetails
-XX:+PrintTenuringDistribution

The following information in the document present here http://docs.sun.com/app/docs/doc/820-4289/abeia?a=view answers my original question!
+The Communications Server uses RMI in the Administration module for monitoring. Garbage cannot be collected in RMI-based distributed applications without occasional local collections, so RMI forces a periodic full collection. Control the frequency of these collections with the property -sun.rmi.dgc.client.gcInterval. For example, - java -Dsun.rmi.dgc.client.gcInterval=3600000 specifies explicit collection once per hour instead of the default rate of once per minute.+
However, this means we are observing anomalous behaviour in our application. The GC logs indicate that FULL GC cycles are NOT happening. Following is what I see in the GC logs when I grep for Full GC
16.892: [Full GC 16.892: [Tenured: 2827K->10193K(1864192K), 0.2641562 secs] 98121K->10193K(2090304K), [Perm : 20479K->20479K(20480K)], 0.2643126 secs]
26.174: [Full GC 26.174: [Tenured: 10193K->12032K(1864192K), 0.2939796 secs] 115023K->12032K(2090304K), [Perm : 24576K->24576K(24576K)], 0.2941415 secs]
31.949: [Full GC 31.949: [Tenured: 12032K->17890K(1864192K), 0.3886069 secs] 185492K->17890K(2090304K), [Perm : 28671K->28671K(28672K)], 0.3887846 secs]
38.411: [Full GC 38.411: [Tenured: 17890K->22908K(1864192K), 0.4984410 secs] 180429K->22908K(2090304K), [Perm : 32768K->32439K(32768K)], 0.4986871 secs]
332.245: [Full GC 332.246: [Tenured[Unloading class sun.reflect.GeneratedMethodAccessor100]
4110.831: [Full GC 4110.831: [Tenured[Unloading class sun.reflect.GeneratedMethodAccessor37]
13818.482: [Full GC 13818.482: [Tenured: 151201K->106589K(1864192K), 1.0052010 secs] 336363K->106589K(2090304K), [Perm : 44799K->44799K(44800K)], 1.0054616 secs]
17622.990: [Full GC 17622.990: [Tenured[Unloading class sun.reflect.GeneratedConstructorAccessor83]The time stamps indicate that Full GC cycles are not happening every minute as per the default value of sun.rmi.dgc.server.gcInterval. I have not changed sun.rmi.dgc.server.gcInterval from its default value (i.e. I have not added an explicit JVM argument to change it from its default value on 1 minute)
Can someone explain why am I not seeing Full GC cycles every minute as expected?
System Properties
Platform: Solaris 10
JDK: 1.5.0_14
Tomcat 5.5.27
Apache 2.2.9

Similar Messages

  • My batterie doesn't work correctly to my Iphone 5. It should work longer, but it's worse. My battery does not hold a full day without much use.

    It should work longer, but it's worse. My battery does not hold a full day without much use.
    When the new version of ISO6 will come ??? A lot of apps don't work correctly too.... :-(
    I'm a little bit disappointed...

    Sounds like your iPhone is locked to your home carrier.
    See if they offer unlocking and if you qualify. Otherwise,
    there is nothing you can do besides rent/but a cheap
    phone for your month in Argentina.

  • JDBC/RMI driver does not support BLOB data type

    WLS 5.1 JDBC/RMI driver does not support BLOB data type. It is said in the Weblogic e-doc.
    I am using DataSource for database connections, meaning JDBC/RMI driver is actually producing
    database connections. I need several columns of BLOB types in my Oracle tables.
    Should I use the "JTS" driver instead in order to have the BLOB types supported ?
    Thanks for help.
    [att1.html]

    The error java.rmi.UnmarshalException seems to indicate you are somehow accessing the data-source over RMI, this seems odd. Are you running TopLink on a client and trying to connect to a WebLogic DataSource on the server?
    Maybe try connecting directly to the DataSource without going through TopLink and see if you encounter the same issue.
    You could also look at the TopLink WebLogic Examples and see how their configuration is different than your own.

  • On resize of report, some data does not display in full

    I have to modify a standard report, so I adjust the main section width from 11 t0 14, and the report width from 132 to 142. I compiled the report and run it. I noticed that some items does not display in full, even the text item header.
    ex:
    prev after modif
    1234.45 123
    Fname Fnam
    Account Accou
    At design mode, I can see that the space is enough for the item to display its full value. Anybody knows to deal with this?
    this report is run thru oracle apps.

    treehuggerinny wrote:
    > This the the search result screen:
    >
    http://67.199.64.241/consumers/srchresult.asp?gender=either&searchword=&zip=&fir
    > stname=&LastName=&go=Search
    >
    > If you choose Herrmann, there should be narratives.
    >
    > Any thoughts would be most appreciated!!!
    I see that you're using ASP, if the fields in question are
    memo fields in Access or nText fields in SQL server, then it's
    possible that the following tech note may help out:
    http://www.webassist.com/media/professional/support/technotes/10009.pdf
    Or you may need to specify all of your columns (not SELECT *
    ) and manually move the memo/ntext fields to the end of the list of
    fields.
    Danilo Celic
    |
    http://blog.extensioneering.com/
    | WebAssist Extensioneer
    | Adobe Community Expert

  • How can I block a number that calls me persistently every single day but only shows 3458  it does not show the full 10 digit

    how can I block a number that calls me persistently every single day but only shows 3458  it does not show the full 10 digit so on your block page will not let me block it!!!  So frustrated and I am not answering it with all the hackers out these days I don't want to give them access to my phone.

        @lottsandy
    I can understand your frustrations and concerns with receiving these unknown calls!  I would love to assist you with preventing these types of calls.
    Sounds as though you're receiving some type of telemarketing call.  Although you're unable to add the shorter numbers to your call block list, you can definitely add your number to the National Do Not Call registry http://1.usa.gov/1whONxp.  I hope this will help prevent those unwanted calls.  Thanks!
    AnthonyTa_VZW
    Follow us on Twitter @VZWSupport

  • Updated to iOS7 on my iPad mini and now anything i type there is a 30 second lag on the letter appearing? Bug already? or does it need a full reset?

    Updated to iOS7 on my iPad mini and now anything i type there is a 30 second lag on the letter appearing? Bug already? or does it need a full reset?

    I have read that if you go to Settings, iCloud, and turn off documents and data that this will help with typing.

  • Does HP tablet support full java ??

    does HP tablet support full jave ??

    saliba wrote:
    thxxxx.
    but which HP tablet model support full java becomes I'm looking for one that support fall jave ?
    I would suggest that you ask that question in the tablet sections, not the Desktop PC section.

  • My doc does not show on full screen, how to project it to full screen?

    The Word document which I want to amend does not show on full screen, how to get the doc to full screen?

    Doc shows Thumb and prints ok, but I jsut cannot see it in full, it is locke in the left hand corner.

  • Does DS invoke Teradata Utilities during processing?

    Does Data Services XI 3.1 invoke Teradata Utilities during processing?
    Edited by: Kim McCormick on Apr 14, 2009 9:10 PM

    Hi,
    I'm using Bulk load to Teradata database with tpt option.
    When I execute my job in Data services, I'm getting error as: "Cannot find the full path for file <tbuild.exe>, due to error <2>. "
    Job console output:
    6024     3872     DATAFLOW     11/4/2010 5:00:09 PM     Data flow <pj_DataFlow> using IN MEMORY Cache.
    6024     4896     BLKLOAD     11/4/2010 5:00:12 PM     Starting Bulk Loader Utility <tbuild> in <Dataflow pj_DataFlow>. Data file is
    6024     4896     BLKLOAD     11/4/2010 5:00:12 PM     <
    .\pipe\pjtarget_Datastore_OF_JIRA_PC2_7_0.dat>.  Command line used is <tbuild.exe -f"C:/Program Files (x86)/Business
    6024     4896     BLKLOAD     11/4/2010 5:00:12 PM     Objects/BusinessObjects Data Services/log/bulkloader/pjtarget_Datastore_OF_JIRA_PC2_7.ctl" OF_JIRA_PC2 -z0>.
    6024     3872     DATAFLOW     11/4/2010 5:00:22 PM     Data flow <pj_DataFlow> is terminated due to error <50616>.
    6024     3872     DATAFLOW     11/4/2010 5:00:22 PM     Process to execute data flow <pj_DataFlow> is completed.
    3596     3324     WORKFLOW     11/4/2010 5:00:27 PM     Work flow <pj_WorkFlow4> is terminated due to an error <50616>.
    3596     3324     JOB     11/4/2010 5:00:27 PM     Job <pj_Job2> is terminated due to error <50616>.
    Please let me know where I'm missing.
    Thanks alot in advance.
    Chintu

  • Muse 7.0 plus Chrome: browser does not show images full browser width

    Hi there,
    I have a serious problem with the latest Muse version: the images which I use as parallax-masks for the background won't work with Chrome, though Safari and Firefox are not showing any problems. The issue: Chrome does either not scale the images to 100% browser-width or does not show them at all, or even shows the images cut into narrow stripes. This problem did not show in the previous version of Muse.
    Of course I already restarted application and computer. System: Mac OSX Mavericks.
    What to do now? The presentation with the client is very soon and this bug makes me very nervous.
    Best,
    bartl

    The problem bartl encountered and reported was fixed by the 7.1 release (November 21st). The problem you're describing sounds different.
    As of yet, I'm unable to reproduce what you're describing. The URL you've provided views the same in all the browsers I've tried, Chrome (32.0.1700.77) and non-Chrome, Mac and Windows.
    That said, looking at your master page I don't think you've created what you intended. At present there are two full screen slideshows on your A-master. On contains no images and the other contains a single image. Both are set to have thumbnails and the thumbnails container is off the right side of the page.
    If you look at your "Master Slideshow" layer in the Layers panel, you'll see the two slideshows. I think you'll want to delete one of them and, given your design, turn off thumbnails for the remaining one.
    Also, if you don't plan to have multiple images in the full screen slideshow, you'd be better off using a background fill image rather than the slideshow widget.
    I remain a bit puzzled how/why you're seeing different behavior in your instance of Chrome. Have you cleared your browser cache?
    On an unrelated note, if you set the page fill and stroke to none and set the browser fill to the most common color from the garage door image, it will avoid the momentary appearance of a white page on a grey background when your home page first loads. Instead the site visitor would see a blank page that's the most common color from the slideshow image, until that image loads.

  • Does iPhoto 6 support full-featured slideshow copying between Macs finally?

    Hi,
    I created long slideshows with the slideshow feature of iPhoto 5.
    It seems that full-featured sharing and copying of that slideshow to another equal equipped Mac is not possible with iPhoto 5. Only the copying of that slideshow as a bare album seems possible with the standard Sharing feature, and with Drag & Drop — while losing all those time consuming optimized transition and timing settings.
    Well, for the "best digital lifestyle applications on the planet" (similar called by Steve), I expect that Apple did work on that issue with iPhoto 6. Have they? Does iPhoto 6 integrate a logic step further now, which allows copying of permanent slideshows and photo books between two Macs?
    Please let me know of advances or tricks on that hardly missed feature.
    I am very curious. Please excuse my English, hope you got the idea.
    Kind regards
    MArQ
    PowerBook G4 1,5 GHz   Mac OS X (10.3.9)   iPhoto 5.0.4
    PowerBook G4 1,5 GHz   Mac OS X (10.3.9)   iPhoto 5.0.4

    Hi MArQ,
    Unfortunately, nothing has changed in that area.
    The only way to do it easily is to create a new library for each slideshow and book specifically. Import the images you are going to use for your book or slideshow and do your book or slideshow in the newly created library.
    This library can then be burned to disc and copied to another computer running the same version of iPhoto. Use the Option key when launching iPhoto to open the newly copied library from the disc.

  • RMI Service does not start

    Hello,
    I am deadly new in this, so I hope this is an easy to answer question ...
    I want to install an EJB application which runs on JBoss. Since JBoss don't need an application.xml file I wrote one and built an .ear file to deploy it into OC4J. Our application uses a fat client which connects to J2EE Container over RMI. Therefore I only installed the EJBs into OC4J. To change as less as possible on our client distribution I used the RMIInitialContextFactory to connect to the server. Problem is, that obviously the RMI server on port 23791 is not started. I got ConnectionExceptions because the connetion is refused. I activated debug logging in OC4J and there are following entries:
    05.11.03 12:17: Error binding to server: com.evermind.server.rmi.OrionRemoteException: Unable to connect to localhost/127.0.0.1:23791: Connection refused: connect;
    The logfile of the instance island in the opmn directory has the entry:
    Auto-deploying vs_ejb.jar (orion-ejb-jar.xml had been updated since the previous deployment)... done.
    Error binding to server: com.evermind.server.rmi.OrionRemoteException: Unable to connect to localhost/127.0.0.1:23791: Connection refused: connect;
    I did change the rmi.xml file and entered a server tag with host="localhost", but it does not work.
    Anyone has an idea ?
    Thanks in advance,
    Frank

    Frank,
    In case you are unaware, there is a stand-alone version of OC4J (which I use) and there is Oracle 9iAS which includes OC4J as one of its components (and there is also an embedded OC4J in "JDeveloper"). Each one behaves slightly differently. Could it be that you were getting confused between the documentation for these products?
    Good Luck,
    Avi.

  • The Full Screen icon does not appear in Firefox 16.0.2 on Mac OS 10.6.8, and command+shift+F does not close the full screen view.

    I accidentally entered full screen view and was unable to exit. The Firefox Help article at https://support.mozilla.org/en-US/kb/how-make-firefox-and-websites-go-full-screen?esab=a&s=full+screen+mac&r=0&as=s says I should see a Full Screen View icon in the upper right-hand corner of the Firefox window. That icon not there. I cannot find it. It does not appear in either the full screen or non-full screen views.
    Further, the Help article's advice to exit full screen view by using the keyboard sequence command+shift+F does not work. I had to unplug my iMac and reboot. Firefox then opened in non-full screen view.
    Ideally, hitting the Escape key would cause Firefox to exit full screen view.

    I forgot one detail. This happened while I was having trouble seeing the mouse cursor (I use Apple's Magic Mouse). It had disappeared briefly. I had to drag the cursor out of the Firefox Window for it to reappear. I accidentally entered full screen view while the cursor was not visible.

  • Magic Mouse does not swipe between full screens.

    Even tho prefrences is checked that one finger swipes between pages and two finger swipe between full screen apps, I am unable to do so.  It only goes between say mail, dashboard and one chrome or say I photo.  What could I be doing wrong?

    I went to Mission Control and set up more desktops, Thanks Captfred.  I'm getting the hang of it, but occ. it gets stuck, ie:  I-Cal goes from page to page.  Any suggestions?  Thanks again Captfred.

  • Why does query do a full table scan?

    I have a simple select query that filters on the last 10 or 11 days of data from a table. In the first case it executes in 1 second. In the second case it is taking 15+ minutes and still not done.
    I can tell that the second query (11 days) is doing a full table scan.
    - Why is this happening? ... I guess some sort of threshold calculation???
    - Is there a way to prevent this? ... or encourage Oracle to play nice.
    I find it confusing from a front end/query perspective to get vastly different performance.
    Jason
    Oracle 10g
    Quest Toad 10.6
    CREATE TABLE delme10 AS
    SELECT *
    FROM ed_visits
    WHERE first_contact_dt >= TRUNC(SYSDATE-10,'D');
    Plan hash value: 915912709
    | Id  | Operation                    | Name              | Rows  | Bytes | Cost (%CPU)| Time     |
    |   0 | CREATE TABLE STATEMENT       |                   |  4799 |  5534K|  4951   (1)| 00:01:00 |
    |   1 |  LOAD AS SELECT              | DELME10           |       |       |            |          |
    |   2 |   TABLE ACCESS BY INDEX ROWID| ED_VISITS         |  4799 |  5534K|  4796   (1)| 00:00:58 |
    |*  3 |    INDEX RANGE SCAN          | NDX_ED_VISITS_020 |  4799 |       |    15   (0)| 00:00:01 |
    Predicate Information (identified by operation id):
       3 - access("FIRST_CONTACT_DT">=TRUNC(SYSDATE@!-10,'fmd'))
    CREATE TABLE delme11 AS
    SELECT *
    FROM ed_visits
    WHERE first_contact_dt >= TRUNC(SYSDATE-11,'D');
    Plan hash value: 1113251513
    | Id  | Operation              | Name      | Rows  | Bytes | Cost (%CPU)| Time     |    TQ  |IN-OUT| PQ Distrib |
    |   0 | CREATE TABLE STATEMENT |           | 25157 |    28M| 14580   (1)| 00:02:55 |        |      |            |
    |   1 |  LOAD AS SELECT        | DELME11   |       |       |            |          |        |      |            |
    |   2 |   PX COORDINATOR       |           |       |       |            |          |        |      |            |
    |   3 |    PX SEND QC (RANDOM) | :TQ10000  | 25157 |    28M| 14530   (1)| 00:02:55 |  Q1,00 | P->S | QC (RAND)  |
    |   4 |     PX BLOCK ITERATOR  |           | 25157 |    28M| 14530   (1)| 00:02:55 |  Q1,00 | PCWC |            |
    |*  5 |      TABLE ACCESS FULL | ED_VISITS | 25157 |    28M| 14530   (1)| 00:02:55 |  Q1,00 | PCWP |            |
    Predicate Information (identified by operation id):
       5 - filter("FIRST_CONTACT_DT">=TRUNC(SYSDATE@!-11,'fmd'))

    Hi Jason,
    I think you're right with kind of "threshold". You can verify CBO costing with event 10053 enabled. There are many possible ways to change this behaviour. Most straightforward would be probably INDEX hint, but you can also change some index-cost related parameters, check histograms, decrease degree of paralellism on table, create stored outline, etc.
    Lukasz

Maybe you are looking for