accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Josh Elser (JIRA)" <>
Subject [jira] [Commented] (ACCUMULO-3346) ActiveScans on monitor page not showing a scan
Date Wed, 19 Nov 2014 20:51:33 GMT


Josh Elser commented on ACCUMULO-3346:

I'm thinking that this might be semi-expected. The SlowIterator does eat Interrupts (which
I think is the root of the problem) so, after the tservers session timeout expires the Session,
the scan is no longer "active". However, if the thread running the actual query (the Iterator
stack) ate the interrupt, it would keep running even though the tserver thinks it isn't.

The same thing might happen at the I/O level if there is something that eats interrupts (which
hints back at the long-standing "long-running scans issue").

I'll see if I can reproduce this again after changing how the iterator handles interrupts.

> ActiveScans on monitor page not showing a scan
> ----------------------------------------------
>                 Key: ACCUMULO-3346
>                 URL:
>             Project: Accumulo
>          Issue Type: Bug
>          Components: monitor, tserver
>    Affects Versions: 1.6.1
>            Reporter: Josh Elser
>             Fix For: 1.6.2, 1.7.0
> Seeing this in the tserver jstack:
> {noformat}
> "User: root Start: 1416425236658 Client: Tablet: 250<<" daemon
prio=5 tid=0x00007f80c692f800 nid=0xe40b waiting on condition [0x000000011d369000]
>    java.lang.Thread.State: TIMED_WAITING (sleeping)
>         at java.lang.Thread.sleep(Native Method)
>         at org.apache.accumulo.core.util.UtilWaitThread.sleep(
>         at
>         at org.apache.accumulo.core.iterators.system.SourceSwitchingIterator.readNext(
>         - locked <0x00000007a66adeb8> (a org.apache.accumulo.core.iterators.system.SourceSwitchingIterator)
>         at
>         - locked <0x00000007a66adeb8> (a org.apache.accumulo.core.iterators.system.SourceSwitchingIterator)
>         at org.apache.accumulo.tserver.Tablet.nextBatch(
>         at org.apache.accumulo.tserver.Tablet.access$3200(
>         at org.apache.accumulo.tserver.Tablet$
>         - locked <0x00000007a4dc4528> (a org.apache.accumulo.tserver.Tablet$Scanner)
>         at org.apache.accumulo.tserver.TabletServer$ThriftClientHandler$
>         at
>         at java.util.concurrent.ThreadPoolExecutor.runWorker(
>         at java.util.concurrent.ThreadPoolExecutor$
>         at
>         at
>         at
>    Locked ownable synchronizers:
>         - <0x00000007a5d362f0> (a java.util.concurrent.ThreadPoolExecutor$Worker)
> {noformat}
> It's a bit confusing. I know there is a scan active because the table in question is
being blocked from deletion (due to a purposefully long-running scan with SlowIterator). However,
it appears incorrect that the monitor is reporting no active scans.

This message was sent by Atlassian JIRA

View raw message