Archive for the ·

AIX

· Category...

AIX version of mktemp command

no comments

If you are looking for the AIX version of mktemp command, which allow safe temporary file creation from shell scripts,  you can actually find the equivalent AIX version in /usr/sbin/lmktemp.

This command comes with fileset bos.rte.lvm.

 

For example,
To create a large file in IBM AIX:

# /usr/sbin/lmktemp myfile 100m
myfile
# du -sm myfile
100.00 myfile

 

 

GPFS cluster tips 1: General commands on AIX

no comments

Terminology:

 

mm ====> Multi Media
NSD ====> Network Shared Disk
mmfsd (1191 is default port) ====> GPFS daemon
Daemon will do I/O and buffer management.
Location of files:
/var/adm/ras/mmfs.log.latest ====> gpfs log file
/usr/lpp/mmfs/bin ====> GPFS command location
/var/mmfs/gen/mmsdrfs ====> Configuration file

Commands:

mmlscluster ====> to list the cluster
 mmgetstate -aLs ====> to view the status of the GPFS cluster
 mmlsconfig ====> Basic configuration information of the GPFS including no.of Filesystems
 lsof -i:1191 ====> to check the daemon port listen state.
 mmlsmgr -c ====> to view the GPFS manager
 mmlsnfd -f <FS-name> -m ====> To check the
 mmlsfs all ====> To check all GPFS file ssytems (lower alphabets are current values)
 mmdf <mount-point> ====> To check the gpfs file system size
 mmdsh ====> To configure trust relation between cluster nodes
 mmlsnsd ====> To
 mmlsdisk <file-system> -d <disk-name> ====> To view the disk information
 mmaddnode ====> to add client node
 mmchnode ====> To change client node name
 mmcrcluster ====> To create gpfs cluster
 /usr/lpp/mmfs/samples ====> location of same files created by installing base filesets
 mmlslicense ====> To view the GPFS license
 mmlsmgr ====> To check the cluster Manager and File Manager
 mmfsadm dump version ====> shows the Version and no.of days cluster is up.
 mmshutdown -a ====> shutdown the GPFS cluser in all nodes.
 mmstartup -a ====> Starts the cluster in all nodes.
 mmfsadm dump config ====> GPFS attributes information
 mmchcluster ====> To change the cluster
 mmlsnsd ====> will show NSD disks
 mmlsnsd -M ====> Show the detailed NSD disks
 mmdf <FS> ====> Show GPFS filesystem details
 mmlsdisk <FS-name> ====> Will shows the disks for that file system
To remove GPFS file system:
 mmumount <file-system> -a ====> To unmount
 mmdelfs <File-system> ====> To remove gpfs filesystem
 mmdelnsd <NSD-name>,<NSD-name> ====> To delete the NSD disk
 mmcrnsd -F /tmp/disk.list ====> To create a nsd disk (/tmp/disk.list file with entries)
 mmcrfs <FS-name> -F </tmp/fs.list> -A yes -T <mount-point> ====>To create File system
 mmadddisk <FS-name> -F </tmp/fs.list> ====> To add new disk to File system
 touch /var/mmfs/etc/ignoreStartupMount ====> To ignore any GPFS filesystem while system is rebooting.
 touch /var/mmfs/etc/ignoreStartupMount.<devicename> ====> to do not mount this FS while system boot

 

 

Monitoring and measuring IO performance for SAN storage from AIX – Part 1

Comments Off

Monitoring SAN storage performance

 

§For random IO, look at read and write service times from

 

# iostat –RDTl <interval> <# intervals>
Disks:                     xfers                                read                                write                                  queue                    time
-------------- -------------------------------- ------------------------------------ ------------------------------------ -------------------------------------- ---------
                 %tm    bps   tps  bread  bwrtn   rps    avg    min    max time fail   wps    avg    min    max time fail    avg    min    max   avg   avg  serv
                 act                                    serv   serv   serv outs              serv   serv   serv outs        time   time   time  wqsz  sqsz qfull
hdisk1           0.9  12.0K   2.5   0.0   12.0K   0.0   0.0    0.0    0.0     0    0   2.5   9.2    0.6   92.9     0    0   3.7    0.0   71.4    0.0   0.0   0.3  15:58:27
hdisk0           0.8  12.1K   2.6 119.4   12.0K   0.0   4.4    0.1   12.1     0    0   2.5   8.7    0.8  107.0     0    0   3.3    0.0   61.1    0.0   0.0   0.3  15:58:27
hdisk2           0.0   0.0    0.0   0.0    0.0    0.0   0.0    0.0    0.0     0    0   0.0   0.0    0.0    0.0     0    0   0.0    0.0    0.0    0.0   0.0   0.0  15:58:27
hdisk3           0.0   0.0    0.0   0.0    0.0    0.0   0.0    0.0    0.0     0    0   0.0   0.0    0.0    0.0     0    0   0.0    0.0    0.0    0.0   0.0   0.0  15:58:27
hdisk4          66.4  58.9M 881.1  58.9M  28.8K 879.1   6.4    0.1  143.5     0    0   2.0   1.8    0.2   27.5     0    0  54.7    0.0    3.6S  48.0   5.0 746.8  15:58:27
hdisk6          66.3  51.1M 797.4  51.0M  24.5K 795.9   7.6    0.1  570.1     0    0   1.5   1.5    0.2   32.9     0    0  51.3    0.0    2.7S  41.0   6.0 678.5  15:58:27
hdisk5          61.9  55.9M 852.9  55.9M  28.5K 850.5   6.0    0.1  120.8     0    0   2.4   1.6    0.1   33.6     0    0  46.1    0.0    3.8S  39.0   5.0 714.7  15:58:27
hdisk7          58.3  55.4M 843.1  55.4M  21.2K 841.9   6.7    0.1  167.6     0    0   1.3   1.3    0.2   20.8     0    0  48.3    0.0    2.1S  40.0   5.0 734.8  15:58:27
hdisk8          42.6  53.5M 729.1  53.5M   3.4K 728.9   5.7    0.1  586.4     0    0   0.2   0.9    0.2    5.9     0    0  54.3    0.0    2.8S  39.0   4.0 687.8  15:58:27
hdisk10         44.1  37.1M 583.0  37.0M  16.9K 582.0   3.7    0.1  467.7     0    0   1.0   1.4    0.2   12.9     0    0  23.1    0.0    1.3S  13.0   2.0 465.0  15:58:27

 

 

 

§Misleading indicators of disk subsystem performance
►%tm_act (percent time active)
●Not meaningful for virtual disks, meaningful for real physical disks
►%iowait
●A measure of CPU idle while there are outstanding IOs
§IOPS, tps, and xfers all refer to the same thing
  • Command topas

# topas –D

or just press D when in topas

 

topas

topas

 

  • Command nmon

# nmon

then press D three times

nmon

nmon

Comments Off