Hello,
We lost the data when the parent regions got reopened. My guess, and
it's only that, is that the regions were essentially empty when they
started up again in these cases. We definitely lost data from the tables.
I've looked through the hdfs and hbase logs and can't find any obvious
difference between a successful split and these failed ones. All steps
show up the same in all cases. After the handled split message that
listed the parent and daughter regions, the next reference is to the
parent regions once again as hbase is started back up after the
failure. No further reference to the daughters is made.
I couldn't cleanly shut several of the regionservers down, so they were
abruptly killed, yes.
HBase version is 0.92.0, and hadoop is 1.0.1.
Thanks.
--Brennon
On 4/11/13 10:58 PM, ramkrishna vasudevan wrote:
When you say that the parent regions got reopened does that mean that you
did not lose any data(any data could not be read). The reason am asking is
if after the parent got split into daughters and the data was written to
daughters and if the daughters related files could not be opened you could
have ended up in not able to read the data.
Some logs could tell us what made the parent to get reopened rather than
daughters. Another thing i would like to ask is was the cluster brought
down abruptly by killing the RS.
Which version of HBase?
Regards
Ram
On Fri, Apr 12, 2013 at 11:20 AM, Brennon Church <[email protected]> wrote:
Hello,
I had an interesting problem come up recently. We have a few thousand
regions across 8 datanode/regionservers. I made a change, increasing the
heap size for hadoop from 128M to 2048M which ended up bringing the cluster
to a complete halt after about 1 hour. I reverted back to 128M and turned
things back on again but didn't realize at the time that I came up with 9
fewer regions than I started. Upon further investigation, I found that all
9 missing regions were from splits that occurred while the cluster was
running after making the heap change and before it came to a halt. There
was a 10th regions (5 splits involved in total) that managed to get
recovered. The really odd thing is that in the case of the other 9
regions, the original parent regions, which as far as I can tell in the
logs were deleted, were re-opened upon restarting things once again. The
daughter regions were gone. Interestingly, I found the orphaned datablocks
still intact, and in at least some cases have been able to extract the data
from them and will hopefully re-add it to the tables.
My question is this. Does anyone know based on the rather muddled
description I've given above, what could have possibly happened here? My
best guess is that the bad state that hdfs was in caused some critical
component of the split process to be missed, which resulted a reference to
the parent regions sticking around and losing the references to the
daughter regions.
Thanks for any insight you can provide.
--Brennon
