HBase - Region in Transition

[#1]

HBase keep having region in transition: 

Regions in Transition

Region State RIT time (ms)

1588230740	hbase:meta,,1.1588230740 state=FAILED_OPEN, ts=Thu Apr 23 12:15:49 ICT 2015 (8924s ago), server=02slave.mabu.com,60020,1429765579823	8924009
Total number of Regions in Transition for more than 60000 milliseconds	1
Total number of Regions in Transition	1

I've try "sudo -u hbase hbase hbck -repair" and also "unassign 'hbase:meta,,1.1588230740'" but still can't fix the problem.

[#2]

Sounds like a bug, check the bugs reported about transitions and compare with the version you are using.

If repair doesn't work, the only solution I see is to do a snapshot, truncate the table and then import the snapshot (maybe try to import into another table before you truncate the main one).

[#3]

Thanks for reply Gonzalo,

I can't do anything with hbase right now (can't disable, drop table, can't event view sample table on hue), everything just stuct, so i just manually delete all table data on HDFS, just keep the default(sample) table, but it still won't work.

[#4]

If restarting the master din't do anything I would say the hbase znode is messed up in zookeeper.

If you have nothing to lose stop hbase, delete the znode in zookeeper, delete the hbase folder in hdfs and start hbase.

[#5]

Thanks Gonzalo,

I've been try to delete the "/hbase" folder ( just move it to /tmp), but when i restart Hbase, the hbase master can't start, as i remember is because of user authentication of "/" belong to HDFS, and i don't want to chown "/" to hbase.

Even after i create a new "/hbase" and chow it to hbase:hbase, the hbase master still won't start unless i move back the old "/hbase"

About znode in zookeeper, i really don't know much about it, i just know my ZooKeeper Znode Parent is "/hbase", do i just delete this folder or i have to delete something elsse ? 

[#6 · Accepted solution]

1. Stop HBase
2. Move your original /hbase back into place
3. Use a zookeeper cli such as "hbase zkcli"[1] and run "rmr /hbase" to delete the HBase znodes
4. Restart HBase. It will recreate the znodes

If Hbase fails to start after this, you can always try the offline Meta repair:
hbase org.apache.hadoop.hbase.util.hbck.OfflineMetaRepair

Also check for inconsistencies after HBase is up.  As the hbase user, run "hbase hbck -details". If there are inconsistencies reported, normally I would use the "ERROR" messages from the hbck output to decide on the best repair method, but since you were willing to start over just run "hbase hbck -repair".

If the above fails, you can always try the offline Meta repair:

hbase org.apache.hadoop.hbase.util.hbck.OfflineMetaRepair

[1] http://hbase.apache.org/book.html#trouble.tools
[2] http://hbase.apache.org/book.html#hbck.in.depth

[#7]

Thanks denloe,

The command "hbase hbck -repair" solve 6 inconsistencies. And the step that delete znodes make this command working, i think, because i've been trying "hbase hbck -repair" before but it just stuck. HBase working fine now, thank you.

[#8]

I have tried your steps, but I still have inconsistancies and hbck -repair does not work. My inconsistancies are with data tables and not with META.

I get the following error message:

INFO util.HBaseFsckRepair: Region still in transition, waiting for it to become assigned

and it eventually times out. I am using CDH 5.4.4 with Hbase 1.0.0. I cannot do anything on Hbase (count, scan etc.). 

[#9]

---

@EugeneM wrote:

I have tried your steps, but I still have inconsistancies and hbck -repair does not work. My inconsistancies are with data tables and not with META.

I get the following error message:

INFO util.HBaseFsckRepair: Region still in transition, waiting for it to become assigned

and it eventually times out. I am using CDH 5.4.4 with Hbase 1.0.0. I cannot do anything on Hbase (count, scan etc.). 

---

In your case, if "hdfs fsck" doesn't fix the files; you are going to have to delete the corrupted hdfs table files.

If you can load the data again, probably the best thing is to delete the /hbase directory in hdfs altogether, restart and load the data again.

[#10]

Thank you Gonzalo. That was exactly what I was not hoping to hear, but seeing that I could not use hbase, I had to do it. So, after deleting all the files/tables with inconsistancies, I got hbase to work again. Thank you for your help.