You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@accumulo.apache.org by "Josh Elser (JIRA)" <ji...@apache.org> on 2015/03/26 00:03:52 UTC

[jira] [Commented] (ACCUMULO-3696) too many trace spans consume memory

    [ https://issues.apache.org/jira/browse/ACCUMULO-3696?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14380983#comment-14380983 ] 

Josh Elser commented on ACCUMULO-3696:
--------------------------------------

{code}
-      if (s.stop - s.start < 1) {
-        synchronized (sendQueue) {
-          sendQueue.remove();
-          sendQueue.notifyAll();
-        }
-        continue;
-      }
{code}

Why the removal on this? Doesn't seem related to the issue you identified and could affect users who do their own tracing.

{code}
+    if (sendQueue.size() > maxQueueSize) {
+      return;
+    }
{code}

Also, any chance at getting some debug message when the Tracer drops spans on the floor? Could also be really frustrating if a user unexpectedly runs into this.

> too many trace spans consume memory
> -----------------------------------
>
>                 Key: ACCUMULO-3696
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-3696
>             Project: Accumulo
>          Issue Type: Bug
>          Components: tserver
>            Reporter: Eric Newton
>            Assignee: Eric Newton
>            Priority: Minor
>          Time Spent: 0.5h
>  Remaining Estimate: 0h
>
> Some users have experienced as many as 100K spans stored in memory, and suspect that it has caused OOM errors.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)