Hi @mike_bronson7 , The thumb rule is to have 1 Gb of heap allocated to 1 Million block. Now, as you have already doubled the heap, I think you can check on improving the garbage collector tunings.   https://community.cloudera.com/t5/Community-Articles/NameNode-Garbage-Collection-Configuration-Best-Practices-and/ta-p/245276   You can try tunings the GC threads to see if that helps. Been said that, these pauses are not very huge to cause any real trouble and we don't see it repeating very frequently ( within mins) and can be expected on a busy Cluster.   You can also check if the threshold to report a JVM pause alert can be adjusted to say 5-10 secs which needs some intervention. ... View more

Hi @hanumanth , The network bandwidth needs to be limited at the OS side using the tools such as traffic control (tc) which can be set on the NIC which carries the ip address for the CDH roles.   For info on this can be found in the Red Hat docs : https://access.redhat.com/solutions/69133 https://access.redhat.com/solutions/1324033     ... View more

Hi @fengsh Yes, The remove script indeed has few bugs using curl which was picked up from AMBARI-18435 but this curl call is not officially supported and as such we don't have an official doc on this.   The other option to remove the older stack with "yum remove <version>" command, specifying the exact version you want to remove should get the job done.   You can also verify the list of packages to be removed using (please be thorough!): # yum list installed | grep -P 'version'   Additionally, please verify the symlinks in the "/usr/hdp/current/" folder are pointing to the correct location.   -- Was your question answered? Please take some time to click on “Accept as Solution” below this post. If you find a reply useful, say thanks by clicking on the thumbs up button. ... View more

Hi @Anlarin ,   It is always suggested to have a homogeneous disk storage across Datanodes. Within datanode, if there are heterogeneous volumes, then when the block replicas are written to new disks on a Round Robin fashion, the disks with less capacity will fill up faster compared to the disks with higher size. If the client is local to Node 2, then it will place the 1st block on that node and it's expected to fill faster.   By choosing "Available Space Policy" the DNs would take into account how much space is available on each volume/disks when deciding where to place a new replica. To achieve writes that are evenly distribution in percentage of capacity on drives, change the choosing policy (dfs.datanode.fsdataset.volume.choosing.policy)to Available Space.   If using Cloudera Manager: Navigate to HDFS > Configuration > DataNode Change DataNode Volume Choosing Policy from Round Robin to Available Space   Click Save Changes Restart the DataNodes   The above property only helps for volumes within Datanode. https://docs.cloudera.com/documentation/enterprise/latest/topics/admin_dn_storage_balancing.html   - Was your question answered? Please take some time to click on “Accept as Solution” below this post. If you find a reply useful, say thanks by clicking on the thumbs up button. ... View more

Hi @isoardi ,   Seeing sockets in TIME_WAIT state is normal and is by design when the socket is getting closed. Unless we see tens of thousands of sockets in TIME_WAIT state which would consume the ephemeral ports on the host , these are fine. It would be the CLOSE_WAIT sockets we need check which indicates the application has not called the close() call on the socket.   You can refer the below RedHat documentation for more info on this and ways to close the TW sockets by reusing them.   https://access.redhat.com/solutions/24154     ... View more

Hi @abdebja ,   You can refer the instructions provided in the below Cloudera Article to mitigate this issue.   https://my.cloudera.com/knowledge/tmp-folder-filling-up-frequently-with-hprof-dump-files?id=340673   - Was your question answered? Please take some time to click on “Accept as Solution” below this post. If you find a reply useful, say thanks by clicking on the thumbs up button. ... View more