Hi. I'm new of this forum, actually this is my first thread.
I am a low-experienced System Administrator on Linux platform.
My English is poor, so I apologize in advance for any wrong way of expressing myself.
Recently I have been put in charge of administer an archiving/backupping system based on TSM 5.3. I'm trying to understand how everything works, but, to be honest, I'm a little bit confused.
What I know about my system:
I have two machines, one works only as dsm client and the other one only as dsm server. Both are connected to a SAN, which is a shared filesystem mounted on /root_data. This is the Archive space. Machines are also connected to a TapeLibrary (IBM 3584) where to migrate and/or backup data.
How I start the System (as by procedure I've inherited):
first of all I execute command “dsmmigfs start” on the two machines, and wait for these processes to start: dsmracalld, dsmmonitord, dsmscoutd, dsmrootd. Actually there is also a “dsmwatchd nodetach” always running.
Then I start dsm-server by command:
and wait for process running:
so I’ll get some TCP port to LISTEN (this is dsm-server):
Then I start dsm-client:
And wait for processes to run:
even here I have some TCP port listening:
This is the dsm.opt on dsm-client:
This is my dsm.sys on dsm-client:
The inclexcl.list on dsm-client:
This is the output of dsmdf command on dsm-client (on the server output is empty):
My schedules:
Since I have dsm-client running just on one machine, is it normal to have those schedules duplicated? I think I should remove di DSMSERVER ones. Do you agree? Also… is it normal dsm-client processes are running just on one machine? Or it should be running on both?
The /root_data is used for 70% and every days dsmautomig hangs, so I restart everything and percentage goes down to 64%.
Please help me to make a good tuning!
If I forgot some important information, just tell me.
I am a low-experienced System Administrator on Linux platform.
My English is poor, so I apologize in advance for any wrong way of expressing myself.
Recently I have been put in charge of administer an archiving/backupping system based on TSM 5.3. I'm trying to understand how everything works, but, to be honest, I'm a little bit confused.
What I know about my system:
I have two machines, one works only as dsm client and the other one only as dsm server. Both are connected to a SAN, which is a shared filesystem mounted on /root_data. This is the Archive space. Machines are also connected to a TapeLibrary (IBM 3584) where to migrate and/or backup data.
How I start the System (as by procedure I've inherited):
first of all I execute command “dsmmigfs start” on the two machines, and wait for these processes to start: dsmracalld, dsmmonitord, dsmscoutd, dsmrootd. Actually there is also a “dsmwatchd nodetach” always running.
Then I start dsm-server by command:
Code:
# dsmserv.rc start
and wait for process running:
Code:
# ps -ef | grep dsm
root 2302 1 1 09:21 pts/1 00:00:00 ./dsmserv QUIET
root 2314 2302 0 09:21 pts/1 00:00:00 ./dsmserv QUIET
root 2315 2314 0 09:21 pts/1 00:00:00 ./dsmserv QUIET
root 2316 2314 0 09:21 pts/1 00:00:00 ./dsmserv QUIET
Code:
# netstat -puntav | grep LISTEN | grep dsm
tcp 0 0 0.0.0.0:1510 0.0.0.0:* LISTEN 2302/dsmserv
tcp 0 0 0.0.0.0:1003 0.0.0.0:* LISTEN 20754/dsmrootd
tcp 0 0 0.0.0.0:1580 0.0.0.0:* LISTEN 2302/dsmserv
tcp 0 0 0.0.0.0:818 0.0.0.0:* LISTEN 20567/dsmrecalld
tcp 0 0 0.0.0.0:1500 0.0.0.0:* LISTEN 2302/dsmserv
Code:
# /etc/init.d/dsmcad start
# /etc/init.d/dsmsched start
# ..../tsm/client/hsm/bin/hsmagent
Code:
# ps -ef | grep dsm
root 2178 1 0 09:25 ? ..../tsm/client/ba/bin/dsmcad -RunAsService
root 2179 2178 0 09:25 ? ..../tsm/client/ba/bin/dsmcad -RunAsService
root 2180 2179 0 09:25 ? ..../tsm/client/ba/bin/dsmcad -RunAsService
root 2181 2179 0 09:25 ? ..../tsm/client/ba/bin/dsmcad -RunAsService
root 2279 1 0 09:25 pts/3 ..../tsm/client/ba/bin/dsmc sch -RunAsService
root 2285 2279 0 09:25 pts/3 ..../tsm/client/ba/bin/dsmc sch -RunAsService
root 2286 2285 0 09:25 pts/3 ..../tsm/client/ba/bin/dsmc sch -RunAsService
Code:
# netstat -puntav | grep LISTEN | grep dsm
tcp 0 0 0.0.0.0:898 0.0.0.0:* LISTEN 8353/dsmrootd
tcp 0 0 0.0.0.0:1581 0.0.0.0:* LISTEN 2178/dsmcad
tcp 0 0 0.0.0.0:36028 0.0.0.0:* LISTEN 2178/dsmcad
tcp 0 0 0.0.0.0:1501 0.0.0.0:* LISTEN 2279/dsmc
tcp 0 0 0.0.0.0:734 0.0.0.0:* LISTEN 7763/dsmrecalld
Code:
SErvername TSM_QUI
tapeprompt no
followsymbolic yes
replace prompt
subdir yes
optionformat standard
restoremigstate no
domain /
domain /root_data
domain /TSM_LOG
Code:
candidatesinterval 1
checkfororphans yes
checkthresholds 5
kernelmessages yes
maxcandprocs 5
maxmigrators 3
maxrecalldaemons 3
maxreconcileproc 3
maxthresholdproc 1
migfileexpiration 3
minmigfilesize 10000
minrecalldaemons 3
reconcileinterval 24
SErvername TSM_QUI
COMMMethod TCPip
TCPPort 1500
TCPServeraddress IP-SERVER
TCPclientaddress IP-CLIENT
schedmode prompted
compression no
memoryefficientbackup yes
tcpwindowsize 64
tcpbuffsize 32
tcpnodelay yes
largecommbuffers no
passwordaccess generate
nodename dsmclient
httpport 1581
tcpclientport 1501
schedlogname ..../tsm/client/ba/log/dsmsched.log
errorlogname ..../tsm/client/ba/log/dsmerror.log
inclexcl ..../tsm/client/ba/bin/inclexcl.list
* dirmc mcdir
schedlogretention 8 D
errorlogretention 8 D
queryschedperiod 1
changingretries 3
commrestartdur 5
commrestartinterv 15
retryperiod 5
maxcmdretries 4
Code:
include /.../* migbck
include /.../*.* migbck
exclude /TSM_DB/*.*
exclude /TSM_DB/*
exclude /TSM_LOG/*.*
exclude /TSM_LOG/*
include /TSM_LOG/tsm_config/* migbck
include /root_data/*.* migbck
include /root_data/* migbck
exclude /root_tmp_data/*.*
exclude /root_tmp_data/*
exclude /root_data/bck/*
exclude /root_data/bck/*.*
include /root_data/mig/.../*.* mig
include /root_data/mig/.../* mig
exclude.backup /root_data/mig/.../*.*
exclude.backup /root_data/mig/.../*
exclude.backup /root_data/mds/.../*.*
exclude.backup /root_data/mds/.../*
include /root_data/migbck/.../*.* migbck
include /root_data/migbck/.../* migbck
exclude.spacemgmt /root_data/bck/.../*.*
exclude.spacemgmt /root_data/bck/.../*
include.backup /root_data/bck/.../*.* bck
include.backup /root_data/bck/.../* bck
exclude /.../*.log migbck
exclude /opt/tivoli/tsm/client/ba/log/* migbck
Code:
# dsmdf
IBM Tivoli Storage Manager
Command Line Space Management Client Interface
Client Version 5, Release 3, Level 4.0
Client date/time: 04/11/12 09:31:32
(c) Copyright by IBM Corporation and other(s) 1990, 2006. All Rights Reserved.
HSM FS Mgrtd Pmgrtd Mgrtd Pmgrtd Unused Free
Filesystem State Size Size Files Files Inodes Size
/root_data a 102.79T 7680K 1403236 15 16.62M 268.18G
Code:
TSM_QUI>query event * * begind=today begint=00:00:00 endd=today endt=23:59:59
04/11/12 19:00:00 BACKUP_GIORNALIERO_ROOT_DATA DSMCLIENT Future
04/11/12 19:00:00 BACKUP_GIORNALIERO_ROOT_DATA DSMSERVER Future
04/11/12 23:30:00 STARTAUTOMIG DSMCLIENT Future
04/11/12 23:30:00 STARTAUTOMIG DSMSERVER Future
The /root_data is used for 70% and every days dsmautomig hangs, so I restart everything and percentage goes down to 64%.
Please help me to make a good tuning!
If I forgot some important information, just tell me.