Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>I suppose you know about <a href="http://www.oracle.com/technetwork/java/javase/gc-tuning-6-140523.html">this reference</a>. <a href="http://www.javabeat.net/2010/08/g1-garbage-collector-in-java-7-0/">This page</a> is also useful. </p> <p>Full GC's occur when tenured objects - those that survive collection in the ephemeral (young) generation - fill up the space allocated for them. When a full GC occurs, any ephemeral marking that was in progress must be aborted. </p> <p>Decreasing the rate at which the tenured generation fills up requires either adding more heap/RAM or fiddling with the division of available memory between tenured and young spaces. The parameters <code>NewSize</code>, <code>MaxNewSize</code> and <code>NewRatio</code> are for the latter. Experiment is the only way to find what will work.</p> <p>Common wisdom is that shifting the ratio to make the tenured generation bigger decreases the number of full collections. In many cases this is true, but not always. There is a condition where many tenured objects are becoming dead shortly after they're tenured. In other words, they should have been collected in the young area, but their life ended a bit too late for that. In this case making the young generation bigger allows these objects to be collected there rather than being tenured. A symptom of this is full collection causing a big decrease in allocated space. </p> <p>That doesn't seem to be your case: <code>2978M-&gt;2498M</code>. Your only out may be to make the heap bigger, buying more memory as needed. Still, nearly all systems that run a long time will have a full collection once in a while. </p>
    singulars
    1. This table or related slice is empty.
    1. POWhat causes the G1 garbage collector in Java 7 to abort its concurrent-mark phase?
      singulars
      1. PTQuestion
    1. PTAnswer
    1. This table or related slice is empty.
    1. USGene
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POWhat causes the G1 garbage collector in Java 7 to abort its concurrent-mark phase?
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUndeletion
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. COCouple of q's: 'Full GC's occur when tenured objects... fill up the space allocated for them'. My tenured space should be 3.5g (4g heap - 512m MaxNewSize). From the young collections just before the full gc, it looks like the amount of data in tenured was 3g. 3.5 > 3, so I would think not full and no GC. Am I misinterpreting that?
      singulars
      1. PO
      1. USsharakan
    2. CO'Still, nearly all systems that run a long time will have a full collection once in a while.' I know this is true with CMS in Java6, although to my mind mainly because of fragmentation. I thought with G1 it would only be the case if the marking threads simply couldn't keep up with garbage production, which with proper tuning should be avoidable. Any reason you think that with G1 you'd have to occasionally run a full gc?
      singulars
      1. PO
      1. USsharakan
    3. COG1 uses many regions. Full collections occur when promotion fails: no empty region is available for compaction. The longer a system runs, the better the chance that every region will have at least one live object.
      singulars
      1. PO
      1. USGene
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload