Hi Udbhav,
Great work on hbase docker deployment was done in
https://issues.apache.org/jira/browse/HBASE-12721 you may start your
journey from there. As for rest of your questions maybe there are some
folks here that were doing similar testing and may give you more info.
Regards
Samir
On Thu,
Hi Sandeep,
What sort of load is on cluster when this is happening ? What logs say in
moment when cluster is this state ? It is strange that whole cluster
is unresponsive i can remember few cases when this is possible:
- "hbase:meta" table is in transition or unavailable
- zookeeper is
In most cases like can try following:
1. stop hbase
2. clean hbase zookeeper data (hbase zkcli --> rmr /hbase)
3. start hbase
Regards
Samir
On Fri, May 13, 2016 at 9:17 AM, Gunnar Tapper
wrote:
> Hi,
>
> I'm doing some development testing with Apache Trafodion running
Hi Amit,
Check regionserver logs, usual suspects, are log running GC and HDFS client
related issues. Check for compaction queue.
Regrads
Samir
On Thu, Mar 31, 2016 at 10:48 AM, Amit Shah wrote:
> Hi,
>
> We have been experimenting hbase (version 1.0) and phoenix (version
Thanks for sharing this Rahul, this is quite useful approach for folks who
have replication already configured and can afford downtime during
migration.
Regards
Samir
On Tue, Mar 1, 2016 at 7:26 PM, rahul gidwani
wrote:
> Hi Samir,
>
> We did this migration without any
We have recently tested similar migration path source: hadoop1/hbase0.94
--> destination: hadoop2:/hbase0.98. Here is steps that we have performed :
1. stop source hbase cluster
2. use distcp to copy /hbase folder to destination cluster
3. run hbase migrate script on destination cluster (on
ecked that blog,helped me to understand what is going on when datanode
> dies.
> But still not sure why it took 15 minutes.
> I will check the server logs again, carefully.
>
> Regards,
> Hironori Ogibayashi
>
> 2016-02-18 22:56 GMT+09:00 Samir Ahmic <ahmic.sa...@gmai
apps/hbase/data/WALs/host7516.mydomain,16020,1455618299902-splitting/host7516.mydomain%2C16020%2C1455618299902..meta.1455690873222.meta,
> length=83, corrupted=false, progress failed=false
> 2016-02-17 15:41:20,482 INFO [RS_LOG_REPLAY_OPS-HOST7518:16020-0]
> coordination.ZkSplitLogWorkerCoord
Hi Hironori,
Based on your logs you have shutdown regionserver that was hosting system
"hbase:meta" table. This is special case since user regions can not be
reassigned to other servers until "hbase:meta" table region is reassigned.
This can partially explain delay in recovery time you have
Hi Anna
What version of hbase you are using ? In most cases regoinserver will be
bind on address which is resolved value of HOSTNAME environment variable.
Try "ping $HOSTNAME" to determine hostname of server then you can add this
line to /etc/hosts:
eth0_ip_address
Hi,
This looks like some WALs splitting operations are not able to finish
causing master to abort. First thing i would try is to stop cluster, clear
zk data and then try to start it again. If that does not solve issue you
probably have case of corrupted WALs files. In that case you will have to
to get all column families for source
> table and create the destination table with them and then set pre-splits.
> Right?But is there any other way to do the same?
>
> Thanks again,
> Sumit
>
>
> From: Samir Ahmic <ahmic.sa...@gmail.com>
> To: user@hbase.apache.org
Hi Sumit,
IMHO snapshots are easiest way to copy tables you need just to steps:
1. create snapshot
2. clone snapshot to another table
What version of hbase you are using? Regarding you question what will
happen in case operation fails in case of cloning snapshot i have never
experience such
want then run CopyTable tool.
Regards
Samir
On Wed, Jan 20, 2016 at 12:24 PM, Ted Yu <yuzhih...@gmail.com> wrote:
> There is also backup / restore (work in progress):
>
> https://issues.apache.org/jira/browse/HBASE-7912
>
> FYI
>
> On Wed, Jan 20, 2016 at 2:1
Based on what i have found in code there are two processes related to
snapshots that are triggered every 5 minutes by default
SnapshotHFileCleaner and SnapshotLogCleaner but none of them can explain
why do you see new snapshots created. Have you been restoring snapshots,
that can explain new
Hi Brian,
Did you have created snapshots on cluster while you were upgrading to new
hbase version ? Based on your logs there are some snapshots operations
running on your cluster. Can you dig your logs to find what is
triggering SnapshotManifestLoader there should some more info just above
this
HI Mukesh,
Did you try to change logging levels in in
$HBASE_CONF_DIR/log4j.properties? You can enable this lines to get debug
info in lour logs:
# Enable this to get detailed connection error/retry logging.
#
Hi,
What sort of load did you have on regionserver when server died? I see
exception happens around compaction and flush operations. Did you check
server's cpu, disks, memory, network... If pause was not caused by GC it
may be that root cause is somewhere else. There is also hdfs exception in
logs
Hi Kristoffer,
It looks like you have some issue with name resolution. Try to remove
incorrect value from reslove.conf (node.dc1.consul) and then restart hbase
cluster.
Regarding issue with region in transition check master log for
"hbase:meta,,1.1588230740"
there should be exception explaining
ehow it feels as if amb2.node.dc1.consul shadows
> > the real host amb2.service.consul.
> >
> > I was looking into the source code and found the configuration
> > 'hbase.regionserver.hostname' - could that be of help here to remove
> > the node.dc1 host?
> >
>
Hi Oussama,
What is the value of "hbase.zookeeper.quorum" in your config ? How many
zookeeper instances you have in your cluster? You can stop zookeeper
process on server with "hbase-daemon.sh stop zookeeper" command but be
careful you mast have majority of zk ensemble up in order to zk cluster
rvice.consul
>
> PING amb1.service.consul (172.17.0.89) 56(84) bytes of data.
> 64 bytes from amb1.node.dc1.consul (172.17.0.89): icmp_seq=1 ttl=64
> time=0.070 ms
>
> ping amb2 from amb2.service.consul
>
> PING amb2.service.consul (172.17.0.90) 56(84) bytes of data.
> 6
Hi,
Can you see snapshot on remote cluster? If you can see snapshot you can use
clone snapshot command from hbase shell to create table.
Regards
Samir
On Oct 14, 2015 6:38 PM, "anil gupta" wrote:
> Hi,
>
> I exported snapshot of a table to remote cluster. Now, i want to
and?
>
> On Wed, Oct 14, 2015 at 11:19 AM, Samir Ahmic <ahmic.sa...@gmail.com>
> wrote:
>
> > Hi,
> > Can you see snapshot on remote cluster? If you can see snapshot you can
> use
> > clone snapshot command from hbase shell to create table.
> > Regards
&
structions to export the snapshot. What
> do you mean by not exporting it to correct directory?
> I am using HDP. Do you mean to that i just need to copy this exported in
> same directory structure as other snapshots?
>
> > On Wed, Oct 14, 2015 at 11:36 AM, Samir Ahmic <a
Hi,
Based on you logs:
2015-09-01 15:35:58,047 INFO [JvmPauseMonitor] util.JvmPauseMonitor:
Detected pause in JVM or host machine *(eg GC): pause of approximately
4954ms*
you had long running GC which cause timeout in communication between
regionserver and zookeeper:
2015-09-01 15:36:04,970 INFO
Hi,
Depending on your hbase version you can get ClusterStatus over api.
https://hbase.apache.org/apidocs/org/apache/hadoop/hbase/ClusterStatus.html
Regards
Samir
On Tue, Sep 1, 2015 at 2:03 PM, James Teng wrote:
> Just as mentioned in the subject, is there any api to
Hi,
It look like you are missing htrace jar in your hadoop classpath. You can
check it with:
$ hadoop classpath | tr : \n | grep htrace
If it is not in classpath you will need to include it in hadop classpth.
HTrace jar is located in $HBASE_HOME/lib.
Regards
Samir
On Tue, Jul 7, 2015 at 1:15
-daemon.sh
start regionserver command.
Thanks!
On Tue, Jul 7, 2015 at 5:19 PM, Samir Ahmic ahmic.sa...@gmail.com wrote:
Hi,
It look like you are missing htrace jar in your hadoop classpath. You can
check it with:
$ hadoop classpath | tr : \n | grep htrace
If it is not in classpath you
Hello,
In RegioServer log file you should have exception above this one you paste
that explains why rs process is exiting. You will probably find root cause
of issue there.
Regards
Samir
On Thu, Mar 26, 2015 at 8:43 AM, ocassano ocass...@staff.voo.be wrote:
Hello everyone,
I’m facing a
Well there should be some more info in logs. You can try to start
RegionServer process with hbase regionserver start commanf that will log
everything on terminal so you can track where is issue.
Regards
Samir
On Thu, Mar 26, 2015 at 10:58 AM, ocassano ocass...@staff.voo.be wrote:
Hi Samir,
This cloud be hdfs related. Did you check your hdfs is running correctly ?
On Thu, Mar 26, 2015 at 2:34 PM, ocassano ocass...@staff.voo.be wrote:
That's what I did to get more logs. But I can't see anything going wrong.
regionserver_start_failed.txt
In this case regionserver process is unable to initialize ShutdownHook
thread. To be honest this something that i have rarely seen, if you are
unable to resolve it i suggest new clean install because sometimes wrong
jars can end up in classpath and is difficult to find root cause especially
if
And this line in your log looks strange:
2015-03-23 16:24:31,327 INFO [regionserver60020] zookeeper.ZooKeeper:
Client environment:java.class.path= ...
did you remove classpath elements or they dont show in log ?
On Thu, Mar 26, 2015 at 3:53 PM, Samir Ahmic ahmic.sa...@gmail.com wrote:
Can
Can you check hbase classpath command maybe you have some strange hadoop
version in your hbase classpath ?
On Thu, Mar 26, 2015 at 3:37 PM, ocassano ocass...@staff.voo.be wrote:
HDFS is up and running fine. It's HEALTHY and not in SAFE MODE.
I watched at HDFS logs while I'm trying to start
Do you use DNS server for name resolution ? Did you setup reverse DNS zone
in your cluster ? I have seen this errors before when there is no reverse
DNS setup. I believe that TableInputFormatBase class requires revers DNS
name resolution.
Regards
Samir
On Wed, Jun 25, 2014 at 10:57 AM, Li Li
?
On Wed, Jun 25, 2014 at 5:08 PM, Samir Ahmic ahmic.sa...@gmail.com
wrote:
Do you use DNS server for name resolution ? Did you setup reverse DNS
zone
in your cluster ? I have seen this errors before when there is no reverse
DNS setup. I believe that TableInputFormatBase class requires revers
Hi Ian,
What hadoop fsck / says ? Maybe you have some corrupted data on your
cluster. Also try using hbase hbck do investigate issue. If you have disk
space issues try adding more data nodes to your cluster. Regarding errors
you have send they are thrown because ProtobufLogWriter is unable to
was clean at the time of the crash and after.
hadoop version 2.4
hbase version 0.98.3
-Ian Brooks
On Tuesday 17 Jun 2014 13:43:45 Samir Ahmic wrote:
Hi Ian,
What hadoop fsck / says ? Maybe you have some corrupted data on your
cluster. Also try using hbase hbck do investigate issue. If you
Hi,
You are definitely experiencing performance degradation cussed by major
compaction. How big is your write load ? Maybe you should consider to
decrease time interval between major compactions if you are writing a lot
data to cluster. Also here is part of hbase book which explains how to
Hi Vimal,
Check 'alter' command in hbase shell. Run help 'alter' in hbase shell there
is very good explanation for command usage.
Regarding impact alter command will change table schema on all regions
belonging to table you are altering.
Regards
Samir
On Fri, May 30, 2014 at 9:07 AM, Vimal
Hi,
This exception:
exception=java.net.SocketTimeoutException: Call to
hadoop77/192.168.1.87:60020 failed because java.net.SocketTimeoutException:
6 millis timeout while waiting for channel to be ready for read. ch :
java.nio.channels.SocketChannel[connected local=/192.168.1.123:33117
-sweep generation total 16252928K, used 11162086K
[0x00041ae0, 0x0007fae0, 0x0007fae0)
concurrent-mark-sweep perm gen total 81072K, used 48660K
[0x0007fae0, 0x0007ffd2c000, 0x0008)
-邮件原件-
发件人: Samir Ahmic [mailto:ahmic.sa...@gmail.com]
发送
Hi, Hansi
Take look at
https://github.com/apache/hbase/blob/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/mapreduce/RowCounter.java
.
You can modified it and and scan.setTimeRange(long minStamp, long maxStamp)
that will give you option to count rows in specific time range.
Anther
Hi David,
You may try this in hbase shell:
scan '.META.', {FILTER=PrefixFilter('table_name')}
this will scan all regions in .META. table with table_name prefix.
Cheers:)
Samir
On Fri, Apr 25, 2014 at 11:49 AM, David Koch ogd...@googlemail.com wrote:
Hi,
How do I get all rows from
Hi David,
Here is quick bash script that will extract encoded region name.
#!/bin/bash
table=$1
echo scan '.META.', {FILTER=\PrefixFilter('$table')\} cmd_file
hbase shell cmd_file | grep $table |awk '{print $1}' | cut -d , -f3 |
cut -d . -f2 |tail -n+2 | uniq
Regards
On Fri, Apr 25, 2014
Hi Vimal,
I have few options how to move data from one hbase cluster to another:
1. You can use org.apache.hadoop.hbase.mapreduce.Export tool to export
tables to HDFS and then you can use hadoop distcp to move data to another
cluster. When data is place on second cluster you can use
cluster.
Is there any other solution which does not require map reduce at all ?
On Fri, Feb 14, 2014 at 2:41 PM, Samir Ahmic ahmic.sa...@gmail.com
wrote:
Hi Vimal,
I have few options how to move data from one hbase cluster to another:
1. You can use
Hi Federico,
You can restart master that should clean rs list.
Regards
2014-02-05 18:15 GMT+01:00 Federico Gaule fga...@despegar.com:
Dear all,
I've introduced a new regionserver to my cluster.
First, there was no rDNS solving names, so it was named with it IP address.
Then i stop the
Hi Amit,
Yes. You can set split policy per table. Here is relevant part of hbase
book:
http://hbase.apache.org/book/regions.arch.html
The policy can set globally through the HBaseConfiguration used or on a per
table basis:
HTableDescriptor myHtd = ...;
Hi,
Which version of hbase you are using ? Did you try fixing with hbase hbck
-fixMeta ?. Also if hbck don't fix this issue for you you can use the
OfflineMetaRepar tool to create META and ROOT regions and tables. You can
find very useful instruction here:
Hi Fernando,
Check logs on RegionServers there should be explanation why you got
NotServingRegionException.
You said that you got error after 10 min of running script ? It could be
that you have some resources issue, memory disk, cpu.
Cheers
On Thu, Jan 16, 2014 at 2:28 PM, Fernando Iwamoto -
Hi Mark,
Base on this exception:
1. 1862 [pool-2-thread-1] WARN
org.apache.hadoop.hbase.zookeeper.RecoverableZooKeeper - Possibly
transient ZooKeeper, quorum=mark-7:2181,
exception=org.apache.zookeeper.KeeperException$ConnectionLossException:
KeeperErrorCode = ConnectionLoss for
Hi Henry,
Are you using automatic major compaction or you turned it off ? From your
logs i can see that issue appeared after compaction is triggered. If few
compactions are run at same time it can severely affect IO performance
leading to errors on RS.(which may be your case). Here is relevant
Hi Vamshi,
What version of hbase you are using ? Regarding *hbase.hregion.max.filesize
*property from hbase-site.xml you should check this from hbase-default.xml :
property
namehbase.hregion.max.filesize/name
value*1073741824*0/value value is in bytes*
description
Maximum
Hi all,
We are using hbase-0.94.6.1 and at moment i'm evaluation Snapshots as
backup solution for moving data between clusters. I'm wondering if someone
have similar experience and what are pros and cons ? Also is Snapshot
future stable enough for this sort of operation ?
Thanks,
Samir
Hi, Asaf
You can run* status 'detailed' | hbase shell* from command line that will
give you detailed status including regions in transition and also you can
use *hbase hbc*k command , and third options is to run *hbase zkcli* and
then *ls /hbase/unassign* that will display encoded names of
Hi, all
I'm evaluating hbase-0.94.6.1 and i have 48 regions on 2 node cluster. I
was restarting on of RSs and after that tried to balance cluster by running
balancer from shell. After running command regions were not distributed to
second RS and i found this line i master log:
2013-04-12
. This is normal behavior for 0.94. I don't know for versions before
that.
Also, are you sure you need 48 tables? And not less tables with more CFs?
JM
2013/4/12 Samir Ahmic ahmic.sa...@gmail.com
Hi, JM
I have 48 tables and as you said it is 1 region per table since i did not
reach
distributed...
unless there's an edge case where when you have only tables with 1 region
it puts them all on the same server :)
Thx,
J-D
On Fri, Apr 12, 2013 at 12:37 PM, Samir Ahmic ahmic.sa...@gmail.com
wrote:
Thanks for explaining Jean-Marc,
We are using 0.90.4 for very long time
HBASE-7060 explains my case, i'm using 0.94.6.1 and looks like issue is
still present.
Thanks for replaying guys
Cheers:)
Hi, Paul
Look like to me that you did not point your java program to correct
HBase configuration. Look at HTablePool constructor at:
http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/client/HTablePool.html#HTablePool()
Maybe you should try it like this:
Configuration conf =
Hi, Kiran
Welcome to beautiful world of HBase transition states :) . When i face
RIT issue this are steps that i use to resolve it:
1. hbase hbck -fixAssignments (this depends on your version of hbase
it can be also just -fix)
If you don't have luck with 1. then you will need manual
Sure thing. I will fill JIRA
Cheers
On Wed, Feb 6, 2013 at 6:20 PM, Ted Yu yuzhih...@gmail.com wrote:
HBASE-5603 has been closed.
Mind filing a new JIRA ?
On Wed, Feb 6, 2013 at 2:47 AM, Samir Ahmic ahmic.sa...@gmail.com wrote:
Hi all,
I have tried rolling-restart.sh and it still hangs
Hi Lu,
I remember that i had similar issue with wrong number of region servers
reported to master. In my case it was issue with reverse name resolution so
i think you should check DNS settings and /etc/hosts.
Try ping -c 2 $HOSTNAME on regionserver that is reported twice
Hi,
We are using hbase-0.90.4, and after one of our region servers went down
master reported that one region is in transition, in CLOSING state:
INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in
transition timed out: access_tokens,,1312452518317.1878726708
state=CLOSING,
66 matches
Mail list logo