01-17-2014 02:54 AM - last edited on 01-17-2014 11:25 AM by jkestelyn
I'm currently using Cloudera Standard 4.8.0 (#50 built by jenkins on 20131125-1015 git: 48c25adb872f94de22b61868e82700217853b60e) and CDH 4.5.0-1.cdh4.5.0.p0.30.
I'm facing a weird problem of HBase Regionserver & Master logs are not getting updated. By default HBase logs configued to
/var/log/hbase path. Current log file size is around 180 MB and configured maximum is 500 MB. Region server maximum back ups is 10.
I tried to restart the hbase cluster using cloudera manager. Still not working.
Please can any one suggest me any configurations missing? why the logs are not getting updated.
If this is not the right forum, kindly can you suggest to which forum should i be asking these.
Please let me know anything else required.
Thanks in advance.
Thanks & Regards,
Sandeep B A
01-19-2014 09:12 PM
@sandeep_ba Can you clarify that your regionserver and master logs in /var/log/hbase were updating just fine until a certain date (what does "ls -lah /var/log/hbase" return?), and then they stopped updating? I think the critical thing to investigate is "what changed on that date?" Do all the log files have the same timestamp (eg. last modified date)? Did you by any chance fill up your local filesytem? (what does "df -kh /var" report?)
01-20-2014 02:12 AM
01-20-2014 02:21 AM
01-20-2014 08:16 PM
01-20-2014 10:37 PM
01-21-2014 05:40 AM
Oh wow! Thanks for following up, I was scratching my head on this one as the logging usually just works. Yes, you've got to be careful when adding custom coprocessors as they can affect HBase in many ways.