Hi,
As far as I understand, a compaction storm occurs when multiple regions of the
same region server perform compaction.
It looks like, from watching hbase UI that within a region server, compaction
is done serially for a specific table.(one region at a time)
So does this mean that a compaction
Hi everyone,
Does anyone know of a ycdb option that will allow the following read request
distribution:
1. 90% of the material read was inserted in the last 24 hours.2. 97% in the
last 2 hours.
The only options i see for request distribution is "latest", "zipfian" and
"uniform"
Is there anyway
Hi everyone,We have experienced a slowdown on writes and we have seen this
error in the rest servers log :
org.apache.hadoop.hbase.regionserver.wal.DamagedWALException: Append
sequenceId=8689, requesting roll of WAL
I have not seen any issues in the region server or master log files.
A few minute
Hi,Since you probably don't have enough space for backup for such a large
table, test it on a small one and rowcount before and after.We work extensively
with mob and there are several issues with it that we had to work around.1.Mob
and ttl - region server crashes after ttl archiver runs.2.Mob a
Thanks for the replies.I understand the difficulty in tracking distributed
scans.Meanwhile I found that "responseTooSlow" queries, do get written to log.
The entry contains scan details(table, start and endrowkey, response size and
processing time)So I was thinking to lower the threshold so that