Member since
09-27-2013
610
Posts
8
Kudos Received
4
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
3757 | 01-28-2014 12:44 PM | |
25316 | 01-21-2014 08:30 AM | |
2526 | 12-12-2013 12:40 PM | |
2803 | 10-02-2013 04:50 PM |
08-19-2014
03:07 PM
More than likely you are hitting HDFS-6621. This is fairly new so it hasn't been fixed in CDH5 as of yet.
... View more
01-28-2014
12:44 PM
1 Kudo
Hi Ashot, are you using RPMs or parcels to manage your Cloudera installation? Could you provide the output from the following commands from all your hosts. rpm -qa | grep cloudera rpm -qa | grep impala -roland
... View more
01-27-2014
06:21 AM
1 Kudo
Hi Ashot, CM 4.8 will only support Impala 1.2 and above so I would verify that you do not have any older versions of Impala (RPMs or parcels) on any of your hosts. We have seen customers update the Impala parcel but have an old RPM installation of Impala still installed. You will also need to add the Catalog Service Role as that is new in Impala 1.2 and you should also restart all of the CM agent services on all hosts in additional to verifying that all the CM agent RPMs were upgraded successully to 4.8 on all hosts. -roland
... View more
01-21-2014
08:30 AM
4 Kudos
Hi Ashok, to increase the file descriptors for the task tracker with a CM controlled cluster you can do the following. Keep in mind that the Alert is only a warning. Verify your current file descriptor ulimit value to your task tracker processes. pid=$(ps faux | grep java | grep tasktracker | awk -F " " '{print $2}');cat /proc/$pid/limits; Perform the following on each node in the cluster. A ulimit of 65536 probably would be enough for most use cases. I just used 262144 as an extreme example. 1. Edit /usr/sbin/cmf-agent and change the ulimit -n setting. 2. Edit /usr/lib64/cmf/service/mapreduce/mapreduce.sh and add ulimit -n 262144 at the top of the script. 3. Push these script changes out to all the slave nodes 4. A few nodes at a time in the same rack, shutdown datanode and tasktracker services on them. 5. On these stopped nodes run a: service cloudera-scm-agent hard_restart 6. Restart the datanode and tasktracker. Verify your file descriptor ulimit value to your task tracker processes after making the change. pid=$(ps faux | grep java | grep tasktracker | awk -F " " '{print $2}');cat /proc/$pid/limits; -roland
... View more
12-12-2013
12:40 PM
Hi Eason, did you hit the back button when adding the host in the wizard? It looks as if CM thinks the host had already been added to the cluster. Are you using templates? If the host had not already been added to the cluster you may want to logout/log back into the CM gui and retry. -roland
... View more
10-02-2013
04:50 PM
1 Kudo
To add to this, Cloudera Manager uses the kadmin interface to generate the service principles. Windows AD does not support the kerberos kadmin interface from my understanding. You will be better off setting up a MIT based Kdc on a linux system and then configuring cross-realm trust with your AD server. -roland
... View more