experiencing issues deploying on hp-cloud

[Charles Butler]

When deploying mapr on hp-cloud i'm getting consistent failures during clustering and creating the maprfs with the following log output.

2014-12-10 17:09:49 INFO cluster-relation-changed Error 3, No such process. Unable to reach mfs. Check for errors in mfs.log.

i did some additional digging and found the following relevant bits in the mapr logs

2014-12-10 17:09:48,677 18399 RunCmd:197 INFO Disk list :
/opt/mapr/server/mrconfig -h 127.0.0.1 -p 5660 disk list
2014-12-10 17:09:48,684 18399 RunCmd:200 ERROR rc=1
2014-12-10 17:09:48,684 18399 RunCmd:201 ERROR 2014-12-10 17:09:48,6836 ERROR Global mrconfig.cc:333 ListDisks rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6837 ERROR Global mrconfig.cc:3694 ListDisk failed Connection reset by peer.(104).
2014-12-10 17:09:48,684 18399 RunCmd:205 INFO Disk list tried. err 1
2014-12-10 17:09:48,684 18399 LogCurrentInfo:962 INFO 2014-12-10 17:09:48,6836 ERROR Global mrconfig.cc:333 ListDisks rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6837 ERROR Global mrconfig.cc:3694 ListDisk failed Connection reset by peer.(104).
2014-12-10 17:09:48,685 18399 RunCmd:197 INFO DiskGroup list :
/opt/mapr/server/mrconfig -h 127.0.0.1 -p 5660 dg list
2014-12-10 17:09:48,691 18399 RunCmd:200 ERROR rc=1
2014-12-10 17:09:48,691 18399 RunCmd:201 ERROR 2014-12-10 17:09:48,6906 ERROR Global mrconfig.cc:478 ListDGs rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6907 ERROR Global mrconfig.cc:3710 ListDiskGroups failed Connection reset by peer.(104).
2014-12-10 17:09:48,691 18399 RunCmd:205 INFO DiskGroup list tried. err 1
2014-12-10 17:09:48,691 18399 LogCurrentInfo:967 INFO 2014-12-10 17:09:48,6906 ERROR Global mrconfig.cc:478 ListDGs rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6907 ERROR Global mrconfig.cc:3710 ListDiskGroups failed Connection reset by peer.(104).
2014-12-10 17:09:48,691 18399 RunCmd:197 INFO sp list :
/opt/mapr/server/mrconfig -h 127.0.0.1 -p 5660 sp list
2014-12-10 17:09:48,698 18399 RunCmd:200 ERROR rc=1
2014-12-10 17:09:48,698 18399 RunCmd:201 ERROR 2014-12-10 17:09:48,6974 ERROR Global mrconfig.cc:536 ListSPs rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6975 ERROR Global mrconfig.cc:3744 ProcessSPList failed Connection reset by peer.(104).
2014-12-10 17:09:48,698 18399 RunCmd:205 INFO sp list tried. err 1
2014-12-10 17:09:48,698 18399 LogCurrentInfo:972 INFO 2014-12-10 17:09:48,6974 ERROR Global mrconfig.cc:536 ListSPs rpc failed Connection reset by peer.(104).
2014-12-10 17:09:48,6975 ERROR Global mrconfig.cc:3744 ProcessSPList failed Connection reset by peer.(104).
2014-12-10 17:09:48,698 18399 LogCurrentInfo:974 INFO
Done capturing debug info

2014-12-10 17:09:48,698 18399 RestoreUdevRules:694 INFO Udev restoration not required.
2014-12-10 17:09:48,698 18399 RunCmd:197 INFO Fileserver stop:
/etc/init.d/mapr-mfs stop
2014-12-10 17:09:49,732 18399 ExitDiskSetup:160 ERROR Error 3, No such process. Unable to reach mfs. Check for errors in mfs.log.[' File "/opt/mapr/server/disksetup", line 1253, in <module>\n RunDiskSetup();\n', ' File "/opt/mapr/server/disksetup", line 1133, in RunDiskSetup\n GetMfsUp();\n', ' File "/opt/mapr/server/disksetup", line 1027, in GetMfsUp\n AbortWithError(errno.ESRCH, msg);\n', ' File "/opt/mapr/server/disksetup", line 190, in AbortWithError\n stack_trace = traceback.format_stack(frame)\n']

[David Tucker (dbtucker)]

Can you confirm that the instances you configured had storage allocated beyond the boot disk ? There should be at least one UNMOUNTED and UNFORMATED storage spindle on each instance that will support the mapr-fileserver service (which for this simple charm is all nodes).

If there is a way to enforce this in the manner that cpu_cores or memory is enforced, I"m happy to add those details to the documentation.