Apache
/
hbase
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HBASE-14368 New TestWALLockup broken by addendum added to parent issue

This commit is contained in:
stack 2015-09-05 08:53:22 -07:00
parent a11f5c55b4
commit bada19bb54
1 changed files with 1 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestWALLockup.java
View File

@ -55,7 +55,6 @@ import org.junit.Test;
import org.junit.experimental.categories.Category;
import org.junit.rules.TestName;
import org.mockito.Mockito;
import org.mockito.exceptions.verification.WantedButNotInvoked;
/**
* Testing for lock up of WAL subsystem.
@ -226,26 +225,13 @@ public class TestWALLockup {
// This append provokes a WAL roll.
dodgyWAL.append(htd, region.getRegionInfo(), key, edit, region.getSequenceId(), true, cells);
boolean exception = false;
Mockito.verify(server, Mockito.atLeast(0)).
abort(Mockito.anyString(), (Throwable)Mockito.anyObject());
try {
dodgyWAL.sync();
} catch (Exception e) {
exception = true;
}
assertTrue("Did not get sync exception", exception);
// An append in the WAL but the sync failed is a server abort condition. That is our
// current semantic. Verify. It takes a while for abort to be called. Just hang here till it
// happens. If it don't we'll timeout the whole test. That is fine.
while (true) {
try {
Mockito.verify(server, Mockito.atLeast(1)).
abort(Mockito.anyString(), (Throwable)Mockito.anyObject());
break;
} catch (WantedButNotInvoked t) {
Threads.sleep(1);
}
}
// Get a memstore flush going too so we have same hung profile as up in the issue over
// in HBASE-14317. Flush hangs trying to get sequenceid because the ringbuffer is held up
// by the zigzaglatch waiting on syncs to come home.