Working papers ...

EMC: Searching for wwn's in initiator groups

2013-10-16T07:27:00.002+02:00

WWN already exists in an initiator group?

symaccess -sid XXX list -type initiator -wwn yourwwn

Symmetrix ID : 000292600XXX

Initiator Group Name
--------------------------------
Exampleinitiator001

Delete wwn from the initiator group.

SPA: Symmetrix ID's are missing in "real time" window.

2012-01-25T07:00:00.001+01:00

stordaemon shutdown all -immediate
stordaemon list -running
stordaemon start storapid
stordaemon start storevntd
stordaemon start storwatchd
stordaemon start storstpd
stordaemon list -running

/etc/init.d/smas stop
/etc/init.d/smas start

TSM trouble - Obtaining Information About Your System's SCSI Devices

2011-11-22T07:37:00.000+01:00

Information about the devices seen by your system is available in the file |/proc/scsi/scsi.This file contains a list of every recognized SCSI device. For example, if a system contains some StorageTek and IBM libraries, a SAN Gateway, and some Quantum DLT drives, the /proc/scsi/scsi file will look similar to this:

Attached devices:|

Host: scsi2 Channel: 00 Id: 00 Lun: 00
Vendor: STK Model: 9738 Rev: 2003
Type: Medium Changer ANSI SCSI revision: 02

Host: scsi2 Channel: 00 Id: 01 Lun: 02
Vendor: PATHLIGHT Model: SAN Gateway Rev: 32aC
Type: Unknown ANSI SCSI revision: 03

etc.....

From this listing, the type of devices, as well as the model name, vendor, and SCSI channel and ID data is available.

EMC - Unable to release device lock 9 on SRDF meta members

2011-11-11T13:13:00.000+01:00

My problem:
symrdf -f devs.txt -sid xxxx -rdfg 2 split fails with error: unable to acquire the Symmetrix device lock
or
symdev -sid xxxx -lock 9 list command shows that some devices in the meta have lock 9 but the head and others do not.
Unable to release the lock 9 on the individual meta members using the command symdev release -sid xxxx -lock 9

How to fix the problem:
Issue the command symdg -force release groupname -lock 9
The response will be Device group 'groupname' locks released for BCV Devices even if there are no SRDF bcv's, and the result will be that the meta members have the lock 9 released.

Check the state of the RDF devices now (symrdf -f devs.txt -sid xxxx -rdfg 2 query) to ensure that they are in a proper state for the split since the previous presence of the lock 9 infers that there was an issue that occurred during an API operation.

EMC CLI - How to manually release lock from device

2011-11-09T08:24:00.000+01:00

Verify that you are experiencing a device lock

To query for external device locks, issue:
symdev -lock list reported a lock on a single device.

To manually release the lock, issue:
symdev -lock # release

!! Use the release lock action only if you believe that the Symmetrix device lock was forgotten and there are NO other operations in progress to the specified Symmetrix devices
.
Be ready to recognize when a device lock being held by a certain application might be allocated as a long duration lock.

Example:
To list a range of Symmetrix devices (000 to 00A) that have a device external lock of 9, enter:
symdev list -sid 870 -lock 9 -RANGE 0:A -v

To release all Symmetrix devices in Symmetrix 870 that have a device external lock of 9, enter:
symdev release -sid 870 -lock 9

*** EMC PREFERRED ****
Release at the device group level if possible:
symdg -lock 9 release

EMC CLI - How to search for locks

2011-11-09T08:20:00.001+01:00

Symmetrix device locks
# symdev -lock list
# symdev -lock -v list

Symmetrix configuration locks
# symcfg -lockn all list
# symcfg -lockn 15 list
# symcfg -lockn 15 -v list

Gatekeeper and Semaphore locks
# symcfg list -semaphores
# symcfg list -v -semaphores

EMC ControlCenter GUI ends Dissolve META with Successful Operation, but no further Operations are then allowed on the dissolved meta members

2011-11-07T09:47:00.000+01:00

Error msg in EGS log:

ConfigSession_23 CFGCHG_ACTN_COMMIT / 5 SYMAPI_CFGSTATE_RUNNING / 1 46 45 0 SYMAPI_C_SUCCESS / 0 0 SYMAPI_C_SRVR_SUCCESS / 0 `The service request was successful` 0
ConfigSession_23 SymConfigChangeSessionEnd ( $handle 1 NULL )
ConfigSession_23 Success
ConfigSession_24 2550 Unable to acquire the Symmetrix device lock

Fix
symdev -sid xxx -force -lock release symdev# (or -range symdevstart:symmdevend)
or
stop / start the Symmetrix Agent that did the ECC config change

How to rescue your partition tables with gpart - and reinstall grub in Red Hat

2011-10-17T08:01:00.000+02:00

Extending disk in vmware.
All the partition tables are missing. Whole disc are unallocated.

start with gpart livecd
Start with gpart /dev/sda (searching for missing partition tables in sda) Write down your cylinder.
Start parted
-unit s
-rescue
--start (cylinder number 'start' in partition1)
--end (cylinder number 'end' in partition1)

Repeat that if you have more than one partition.

You will now find sda1 and sda2 in /dev/*.

Sda1 = boot (ok)
sda2 = / (ok)
sda3 = SWAP-sda3 (I recreate a new swap disk with gpart, it is not necessary but it is far easier because sda3 has many links around the configuration files in my system.)

mount sda2 to /mnt/sda2/
mount sda1 to /mnt/sda2/boot/

go too /mnt/sda2/sbin/
start: grub-installation /dev/sda2
installing grub.

It perfectly works.

In some cases you also need to reconfigure the network card.
nano /etc/sysconfig/network-scripts/ifcg-eth0 .

Scan HBA for new devices - PowerPath

2011-10-13T07:39:00.000+02:00

The host bus adapter driver in the 2.6 kernel exports the scan
function to the /sys directory which can be used to rescan the SCSI
devices on that interface. The scan function is available as follows:
# cd /sys/class/scsi_host/hostX/
# ls -al scan
# echo ‘- - -’ > scan
"X" will be the HBA number which is zoned to see the disk.
The three dash marks refer to channel, target, and LUN numbers.

Commit your change
#powermt config

Find your new logical device and write down the Pseudo name.
#powermt display dev=all

Find your new disk and pseudo name in
#fdisk -l

Run fdisk command which used to create and manipulate partition tables.
#fdisk
n - 1 - t (83 for linux system)
You should find pseudname and pseudname1 in /dev/
(in my system emcpowerb - emcpowerb1)

Format the new partition with
#mkfs.ext3 /dev/pseudoname1/

Mount the new disk in a directory and use it.

All server sessions are currently in use

2011-09-15T11:37:00.001+02:00

ANS1351E Session rejected: All server sessions are currently in use

Problem solving example:
tsm> setopt maxsession 750
tsm> set maxschedsession 100

This parameter specifies the maximum number of simultaneous client sessions that can connect with the TSM server

How long since the last time the tape cartridge have been used

2011-09-10T12:50:00.000+02:00

A tape cartridge is a mechanical product requiring physical maintenance once in a while. All you need is run a query for how long since the last time the tape cartridge have been used.

select volume_name,date(last_write_date) as write_date,date(last_read_date) as read_date from volumes where stgpool_name='LTO4POOL' order by read_date asc

You can check the tape cartridge for inconsistencies.

audit volume volumename fix=no

TSM for Linux - nightly routine full backup for CentOS

2011-09-06T17:26:00.000+02:00

Script created in /etc/cron.daily/

[root@name cron.daily]# cat dsmc.sh
#!/bin/bash
cd /var/log
/opt/tivoli/tsm/client/ba/bin/dsmc incremental / -subdir=yes | mail -s "backup report " mail@mail.com

Examples - restore:dsmc restore -filelist=/home/dir2/restorelist.txt /home/NewRestoreLocation/
dsmc restore /home/monnett/h1.doc -latest
dsmc restore "/user/project/*"-pick -inactive
dsmc restore /home/monnett/h1.doc /home/newdoc/h2.doc
dsmc restore -pitd=9/10/2011 -pitt=13:23:20 /home/mydir/
dsmc restore "/home/projectA/*.bak" /home/projectB/

Examples - backup:
dsmc incremental / -subdir=yes
dsmc incremental -domain="/fs1 /fs2"
dsmc selective /home/proj/ -subdir=yes
dsmc selective /usr/dir1/sub1/ -subdir=yes -snapshotroot=/snapshot/day1