Troubleshooting Guide, Release 6.0.1
Disk Replacement
Downloads: This chapterpdf (PDF - 417.0KB) The complete bookPDF (PDF - 11.41MB) | Feedback

Disk Replacement

Table Of Contents

Disk Replacement

Introduction

System Back Up Procedure

Call Agent/Feature Server Back Up

Element Management System/Bulk Data Management System Back Up

Call Agent/Feature Server or Element Management System Disk 0 Replacement

Call Agent/Feature Server or Element Management System Disk 1 Replacement


Disk Replacement


Revised: July 22, 2009, OL-15636-07

Introduction

This chapter describes the process of replacing a defective Cisco BTS 10200 Softswitch disk. Each Cisco BTS 10200 system element contains two hard disks (disk 0 and disk 1) that are mirrored using Sun Solaris Disk Suite. The disk mirroring application must be disabled between the two disks in order to replace the defective disk.

Before starting the disk replacement procedures, the following procedures must be completed:

Identify the disk that needs to be replaced by viewing the /var/adm/messages

Perform a full system back up.

The part number of the replacement disk must be the same as the part number of the defective disk, or the size of the replacement disk must be equal to or larger than the mirrored good disk.

This chapter contains the following sections:

System Back Up Procedure

Call Agent/Feature Server or Element Management System Disk 0 Replacement

Call Agent/Feature Server or Element Management System Disk 1 Replacement

System Back Up Procedure

The purpose of this section is to describe the needed steps to create a back up using the Flash Archive on Cisco BTS 10200 system. The procedure is performed before and after any software upgrade or for routine maintenance purposes.


Note It is recommended that this procedure be performed during a maintenance window or when traffic volume is low.


Before you start the back up procedure, obtain the following information:

NFS server hostname or IP address

Shared directory from NFS server

Root user access


Note Provisioning is not allowed during procedure, so that a stable backup is preserved.


Call Agent/Feature Server Back Up

Perform the following steps on the standby CA/FS system:


Step 1 Log in as root on secondary CA/FS.

Step 2 Verify that all platforms are in standby mode.

<hostname># nodestat

Step 3 Remove all unnecessary files or directories such as /opt/Build, application tar files, and so on.

Step 4 Mount NFS server to /mnt directory.

<hostname># mount <nfs server ip or hostname>:/<share dire> /mnt

Step 5 Stop all platforms.

<hostname># platform stop all

Step 6 Save all platforms data directory (shared memory) to nfs server.

<hostname># tar -cf - /opt/OptiCall/CAxxx/bin/data |gzip -fast - > 
/mnt/data.<hostname>.CA.gz
<hostname># tar -cf - /opt/OptiCall/FSAINxxx/bin/data |gzip -fast - > 
/mnt/data.<hostname>.FSAIN.gz
<hostname># tar -cf /opt/OptiCall/FSPTCxxx/bin/data |gzip -fast - > 
/mnt/data.<hostname>.FSPTC.gz


Note Where xxx is the instance number.


Step 7 Start all platforms.

<hostname># platform start -nocopy

Step 8 Verify all platforms are up on standby.

<hostname># nodestat

Step 9 Create excluded directories file for flash archive.

<hostname># vi /tmp/excluded_dir/opt/OptiCall/CAxxx/bin/data
<hostname># vi /tmp/excluded_dir/opt/OptiCall/CAxxx/bin/logs
<hostname># vi /tmp/excluded_dir/opt/OptiCall/FSAINxxx/bin/data
<hostname># vi /tmp/excluded_dir/opt/OptiCall/FSAINxxx/bin/logs
<hostname># vi /tmp/excluded_dir/opt/OptiCall/FSPTCxxx/bin/data
<hostname># vi /tmp/excluded_dir/opt/OptiCall/FSPTCxxxx/bin/logs
<hostname># vi /tmp/excluded_dir/opt/Build


Note Where xxx is the instance number.


Step 10 Back up the system.

<hostname># mv /bin/date /bin/date.archive
<hostname># mv /bin/.date /bin/date
<hostname># flarcreate -n <hostname> -I -X /tmp/excluded_dir -c /mnt/<hostname>.archive
<hostname># mv /bin/date /bin/.date
<hostname># mv /bin/date.archive /bin/date

Step 11 Un-mount NFS server.

<hostname># umount /mnt

Step 12 Switch over all platforms.

Perform the following commands from the Active-EMS.

<hostname># ssh optiuser@<hostname>
CLI> control feature-server id=FSAINxxx; target-state=standby-active;
CLI> control feature-server id=FSPTCxxx; target-state=standby-active;
CLI> control call-agent id=CAxxx; target-state=standby-active;


Note Where xxx is the instance number of each platform.


Step 13 Repeat the same steps for the primary CA/FS.


Element Management System/Bulk Data Management System Back Up

Perform the following steps on standby EMS/BDMS system:


Step 1 Remove all unnecessary files or directories such as /opt/Build, application tar files, and so on.

Step 2 Mount NFS server to /mnt directory.

<hostname># mount <nfs server ip or hostname>:/<share dire> /mnt

Step 3 Stop all platforms.

<hostname># platform stop all

Step 4 Save Oracle database and msql directories.

<hostname># tar -cf - /data1/oradata |gzip -fast - >/mnt/oradata.<hostname>.gz

Step 5 Create excluded directories file for flash archive.

<hostname># vi /tmp/excluded_dir/data1/oradata
<hostname># vi /tmp/excluded_dir/opt/data/oradata
<hostname># vi /tmp/excluded_dir/opt/Build

Step 6 Start all platforms.

<hostname># platform start -nocopy

Step 7 Verify all platforms are up on standby.

<hostname># nodestat

Step 8 Back up the system.

<hostname># mv /bin/date /bin/date.archive
<hostname># mv /bin/.date /bin/date
<hostname># flarcreate -n <hostname> -I -X /tmp/excluded_dir -c /mnt/<hostname>.archive
<hostname># mv /bin/date /bin/.date
<hostname># mv /bin/date.archive /bin/date

Step 9 Unmount NFS server.

<hostname># umount /mnt

Step 10 Switch over all platforms.

Perform the following commands from the active EMS.

<hostname># ssh optiuser@<hostname>
CLI> control bdms id=BDMS01; target-state=standby-active;
CLI> control element-manager id=EM01; target-state=standby-active;

Step 11 Repeat the same steps for Primary EMS/BDMS.


Call Agent/Feature Server or Element Management System Disk 0 Replacement


Note The letter "x" in cxt0d0 or cxt1d0 throughout the procedure refers to the SCSI controller ID of the system (either 0 or 1). To find out the controller ID, use the format command.



Step 1 Log in as root.

Step 2 Verify application status.

<hostname>#nodestat


Note If all platforms on this node are ACTIVE, then you need to switch over to STANDBY side using the CLI command.


Step 3 Stop all Cisco BTS 10200 platforms.

<hostname>#platform stop all

Step 4 Perform the following commands to detach all submirror metadevice.

<hostname>#metadetach -f d2 d0
<hostname>#metadetach -f d14 d12
<hostname>#metadetach -f d11 d9
<hostname>#metadetach -f d5 d3
<hostname>#metadetach -f d8 d6

Step 5 Perform the following commands to clear all submirror metadevice.

<hostname>#metaclear d0
<hostname>#metaclear d12
<hostname>#metaclear d9
<hostname>#metaclear d3
<hostname>#metaclear d6

Step 6 Delete database replica on defective disk.

<hostname>#metadb -d -f /dev/dsk/cxt0d0s4

Step 7 Verify that there is no existing metadevice on defective disk.

<hostname>#metastat -p |grep cxt0d0

Step 8 Use the cfgadm command to determine the defective disk parameter.

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
configured
unknown
c1::dsk/c1t1d0
disk
connected
configured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok

Step 9 Take disk 0 off line.

<hostname>#cfgadm -c unconfigure c1::dsk/cxt0d0

Example:

Secca38#cfgadm -c unconfigure c1::dsk/c1t0d0

Step 10 Verify that Disk 0 is off line.

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
unconfigured
unknown
c1::dsk/c1t1d0
disk
connected
configured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok

secca38#tail /var/adm/messages

Nov  2 15:53:25 secca38 genunix: [ID 408114 kern.info] /pci@1f,700000/scsi@2/sd@0,0 (sd0) 
offline

Step 11 Remove the defective disk and insert the new replacement disk into slot 0.

Step 12 Use the cfgadm command to bring the new Disk 0 on line.

<hostname># cfgadm -c configure c1::dsk/cxt0d0

Example:

Secca38#cfgadm -c configure c1::dsk/c1t0d0

Step 13 Verify that the new Disk 0 is on line.

<hostname>#cfgadm -al

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
configured
unknown
c1::dsk/c1t1d0
disk
connected
configured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok

secca38#tail /var/adm/messages

Nov  2 15:53:25 secca38 genunix: [ID 408114 kern.info] /pci@1f,700000/scsi@2/sd@0,0 (sd0) 
online

Step 14 Use the format and prtvtoc commands to partition the new disk

Disk layout:

Slice 0 /(root)2000MB
Slice 1 /var 5000MB
Slice 3 swap 4000MB
Slice 4 unassigned 24MB
Slice 5 /opt The rest of the disk space
Slice 6 unassigned 2000MB

Step 15 Create the database replica on the new disk.

<hostname>#metadb -a -f -c 3 /dev/dsk/cxt0d0s4

Step 16 Perform the following commands to create submirror metadevice.

<hostname>#metainit -f d0 1 1 /dev/dsk/cxt0d0s0
<hostname>#metainit -f d12 1 1 /dev/dsk/cxt0d0s6
<hostname>#metainit -f d9 1 1 /dev/dsk/cxt0d0s5
<hostname>#metainit -f d3 1 1 /dev/dsk/cxt0d0s1
<hostname>#metainit -f d6 1 1 /dev/dsk/cxt0d0s3 

Step 17 Install the boot block on the new disk.

<hostname>#installboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/cxt0d0s0

Step 18 Sync Disk 0 (new disk) with Disk 1.

<hostname># metattach d2 d0
<hostname>#metattach d14 d12
<hostname>#metattach d11 d9
<hostname>#metattach d5 d3
<hostname>#metattach d8 d6


Note The disk sync process with take awhile depending on the disk size. To find out the percentage completion of this process, you can run the metastat|grep % command.


Step 19 Start all Cisco BTS 10200 platforms.

<hostname>#platform start -nocopy


Call Agent/Feature Server or Element Management System Disk 1 Replacement


Note The letter "x" in cxt0d0 or cxt1d0 throughout the procedure refers to the SCSI controller ID of the system (either 0 or 1). To find out the controller ID, use the format command.



Step 1 Log in as root.

Step 2 Verify application status.

<hostname>#nodestat


Note If all platforms on this node are ACTIVE, then you need to switch over to STANDBY side using the CLI command.


Step 3 Stop all Cisco BTS 10200 platforms.

<hostname>#platform stop all

Step 4 Perform the following commands to detach all submirror metadevice.

<hostname>#metadetach -f d2 d1
<hostname>#metadetach -f d14 d13
<hostname>#metadetach -f d11 d10
<hostname>#metadetach -f d5 d4
<hostname>#metadetach -f d8 d7

Step 5 Perform the following commands to clear all submirror metadevice.

<hostname>#metaclear d1
<hostname>#metaclear d13
<hostname>#metaclear d10
<hostname>#metaclear d4
<hostname>#metaclear d7

Step 6 Delete database replica on defective disk.

<hostname>#metadb -d -f /dev/dsk/cxt1d0s4

Step 7 Verify that there is no existing metadevice on defective disk.

<hostname>#metastat -p |grep cxt1d0

Step 8 Use the cfgadm command to determine the defective disk parameter.

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
configured
unknown
c1::dsk/c1t1d0
disk
connected
configured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok

Step 9 Take disk 1 off line.

<hostname>#cfgadm -c unconfigure c1::dsk/cxt1d0

Example:

Secca38#cfgadm -c unconfigure c1::dsk/c1t1d0

Step 10 Verify that Disk 1 is off line.

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
configured
unknown
c1::dsk/c1t1d0
disk
connected
unconfigured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok

secca38#tail /var/adm/messages

Nov  2 15:53:25 secca38 genunix: [ID 408114 kern.info] /pci@1f,700000/scsi@2/sd@0,1 (sd1) 
offline

Step 11 Remove the defective disk and insert the new replacement disk into slot 1.

Step 12 Use the cfgadm command to bring the new Disk 1 on line.

<hostname># cfgadm -c configure c1::dsk/cxt1d0

Example:

Secca38#cfgadm -c configure c1::dsk/c1t1d0

Step 13 Verify that the new Disk 0 is on line.

<hostname>#cfgadm -al

Example:

secca38#cfgadm -al

Ap_Id
Type
Receptacle
Occupant
Condition
c0
scsi-bus
connected
configured
unknown
c0::dsk/c0t0d0
CD-ROM
connected
configured
unknown
c1
scsi-bus
connected
configured
unknown
c1::dsk/c1t0d0
disk
connected
configured
unknown
c1::dsk/c1t1d0
disk
connected
configured
unknown
c2
scsi-bus
connected
unconfigured
unknown
usb0/1
unknown
empty
unconfigured
ok
usb0/2
unknown
empty
unconfigured
ok


secca38#tail /var/adm/messages

Nov  2 15:53:25 secca38 genunix: [ID 408114 kern.info] /pci@1f,700000/scsi@2/sd@0,1 (sd1) 
online

Step 14 Use the format and prtvtoc commands to partition the new disk.

Disk layout:

Slice 0 /(root)2000MB
Slice 1 /var 5000MB
Slice 3 swap 4000MB
Slice 4 unassigned 24MB
Slice 5 /opt The rest of the disk space
Slice 6 unassigned 2000MB

Step 15 Create the database replica on the new disk.

<hostname>#metadb -a -f -c 3 /dev/dsk/cxt1d0s4

Step 16 Perform the following commands to create submirror metadevice.

<hostname>#metainit -f d0 1 1 /dev/dsk/cxt1d0s0
<hostname>#metainit -f d12 1 1 /dev/dsk/cxt1d0s6
<hostname>#metainit -f d9 1 1 /dev/dsk/cxt1d0s5
<hostname>#metainit -f d3 1 1 /dev/dsk/cxt1d0s1
<hostname>#metainit -f d6 1 1 /dev/dsk/cxt1d0s3

Step 17 Install the boot block on the new disk.

<hostname>#installboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/cxt1d0s0

Step 18 Sync Disk 1 (new disk) with Disk 0.

<hostname># metattach d2 d1
<hostname>#metattach d14 d13
<hostname>#metattach d11 d10
<hostname>#metattach d5 d4
<hostname>#metattach d8 d7


Note The disk sync process with take awhile depending on the disk size. To find out the percentage completion of this process, you can run the metastat|grep % command.


Step 19 Start all Cisco BTS 10200 platforms.

<hostname>#platform start -nocopy