AnsweredAssumed Answered

MapR container issues

Question asked by altoh on Mar 20, 2017
Latest reply on Apr 10, 2017 by cathy

Hi Folks.. 


I have some issues on hand that we are unable to resolve and would need some help on it.


2 disks were replaced in a SP last week.  Following the disks replacement, the following were observed


Issue 1:

The exported mount point from MapR db is very very slow from the NFS client.  Restarting the mapr-nfsserver does not help.  The latency in accessing the exported mount point from the client can be as slow as 10seconds.


Issue 2:

The new disks are not being written into even after 1 week.  The FS of these 2 disks are "MapR-FS" but disk utilisation is 0%.


Issue 3:-

We are getting container issues from what we observed in /opt/mapr/logs/mfs.log

2017-03-20 07:26:01,0667 ERROR x.x.0.0:0 ResyncContainerFrom received error in doresync resp srccid 5538 replicacid 5538 remotenode err:11

2017-03-20 07:44:57,1432 ERROR x.x.1.96:5660 GetAttr 30376.2830.233092 : GetContainer failed 19
2017-03-20 07:44:57,1432 ERROR x.x.1.96:5660 container 30376 is stale, and hasn't been resynced yet


[root@SAW02 server]# maprcli dump containerinfo -ids 30376 -json
*** warn: MAPR_SUBNETS set to
"OwnedSizeMB":"14.24 GB",
"SharedSizeMB":"0 MB",
"LogicalSizeMB":"14.24 GB",
"TotalSizeMB":"14.24 GB",
"Mtime":"Tue Mar 07 03:33:56 UTC 2017",


Version of MapR is pretty old, we are at  We were trying to upgrade to the next version but such things happen before we can proceed.