Lucent Technologies CentreVu Call Management System Release 3 Version 8 Guide

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
State Database Replicas75
State Database Replicas 3
Setting up replicas31. Verify that the /etc/opt/SUNWmd/md.tab file is correct (see 
“Examples of valid system files” on page 79
).
2. Enter the metadb
 command:
Replacing replicas3Sometimes replicas may have to be replaced. A replica can become 
corrupted, for example, and in that case needs to be removed and then 
recreated. The procedure to do that is as follows:
1. Remove all existing replicas with the following command:
Do not reboot the system after removing replicas! With replicas gone, 
the system may not boot!
2. Recreate two replicas on the same partition:# metadb -a -c4 -f /dev/dsk/c0t0d0s1
# metadb -a -c3 -f /dev/dsk/c0t1d0s1
# metadb -d -f /dev/dsk/c0t0d0s1
# metadb -d -c4 -f /dev/dsk/c0t0d0s1
!WARNING: 
						

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Partitioning disks with the format command76
Partitioning disks with the format command3
To use the format command to partition a disk, do the following steps:
1. Enter:# format
A menu of disks displays. You are asked to select one.
2. Enter the menu number of the disk to be partitioned. A list of 
commands displays, followed by the format> 
prompt.
3. Enter:format> partition
The partition> 
prompt displays.
4. Enter:partition> print
The partition table displays and partition>
 returns.
5. Enter the number of the partition to be configured. The system 
begins a series of prompts for partitioning information. Answer 
according to the CMS version, disk type (boot or nonboot), and disk 
size you are partitioning.
Partitioning (Note: Never change the default size of the 
overlap slice!)
CMS R3V8
4.2-GB DiskCMS R3V8
9.1-GB Disk 
SCSI (E3000) 
and FC-AL (E 
3500)
Disk 
TypePart Tag FlagSize
(cyl.)Starting
CylinderSize
(cyl.)Starting 
Cylinder
Boot 0boot disk 1: / or rootboot disk 2: unassignedwm 1023 0 616 0
1 unassigned wm 7 1023 7 616
2overlap 
or backup wm 3880 0 4924 0
3 unassigned wm 1879 1030 3716 623
4 swap wm 971 2909 585 4339
5-7unassignedwm0000
Nonboot0 unassignedwm2020
1 unassigned wm 3878 2 4922 2
2overlap 
or backup wm 3880 0 4924 0
3-7unassignedwm0000 
						

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Partitioning disks with the format command77
6. Enter:partition> print
The partition table displays and the partition>
 prompt returns. 
Compare the display to the 
Partitioning table to make certain you 
have repartitioned the disk properly. If not, repeat step 5.
7. Enter:partition> label
The system responds Ready to label disk, continue?
.
8. Enter:y
9. The partition>
 prompt returns.
10. Enter:partition> q
The format> 
prompt displays.
11. To select another disk to partition, enter:format> disk
and return to item 2.
To quit, enter: format> q
and continue with the next step.
12. When the system prompt returns, verify the partitioning.
You can verify the exact partitioning of the new disk with an 
olds -mirrored -check_disks 
command. For example:
# olds -mirrored -check_disks c0t2d0
scsi=c1
number of external scsi controllers with disks 
is = 1
number of disks is = 6
valid disks are c0t2d0
disk:c0t2d0 is partitioned ok
# 
						

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
How to tell whether your mirrored system is standard78
How to tell whether your mirrored system is standard3
To tell whether mirroring was installed as a “standard” or “nonstandard” 
configuration, you can do one of two things. Either:
a. Enter the command:# metastat d11 d12
b. Edit the 
/etc/opt/SUNWmd/md.tab file
Examine the command output or the contents of the file. You are looking 
for the d11 and d12 metadevices and their constituent disk drives.
Standard mirroring—mirroring that was installed at the factory or 
upgraded to a standard factory configuration—will have the structure 
indicated in the following table:.
Root Metadevice Device Names
Metadevice Device Name
 metadevice d11  c0t0d0
 metadevice d12  c0t1d0 (E3000)
c1t4d0 (E3500) 
						

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Examples of valid system files79
Examples of valid system files3
Va l i d  vfstab files3
Va li d  vfstab file for a non mirrored R3V8 CMS system:
___________________________________________________________________
#device device mount FS fsck mount mount
#to mount to fsck point type pass at boot options
#
#/dev/dsk/c1d0s2 /dev/rdsk/c1d0s2 /usr ufs 1 yes -
fd - /dev/fd fd - no -
/proc - /proc proc - no -
/dev/dsk/c0t0d0s4 - - swap - no -
/dev/dsk/c0t0d0s0 /dev/rdsk/c0t0d0s0 / ufs 1 no -
swap - /tmp tmpfs - yes -
/dev/md/dsk/d19 /dev/md/rdsk/d19 /cms ufs 2 yes
Va li d  vstab file for an R3V8 CMS system with mirrored disks:
___________________________________________________________________
#device device mount FS fsck mount mount
#to mount to fsck point type pass at boot options
#
fd - /dev/fd fd - no -
/proc - /proc proc - no -
/dev/md/dsk/d17 - - swap - no -
/dev/md/dsk/d13 /dev/md/rdsk/d13 / ufs 1 no -
/dev/md/dsk/d21 /dev/md/rdsk/d21 /cms ufs 2 yes - 
						

							  Maintaining Mirrored Systems  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Examples of valid system files80
Va l i d  md.tab files3
Va li d  md.tab file for a 3 disk non mirrored R3V8 CMS system:
____________________________________________________
#state database replicas
mddb00 c0t0d0s1
mddb01 c0t1d0s0
#/cms
d19 3 1 /dev/rdsk/c0t0d0s3 1 /dev/rdsk/c0t1d0s1 1 /dev/rdsk/c1t0d0s1
d21 -m d19
____________________________________________________
Va li d  md.tab file for a 4 disk mirrored R3V8 CMS system:
____________________________________________________
#state database replicas
mddb00 c0t0d0s1
mddb01 c1t4d0s1
mddb02 c0t1d0s0
mddb03 c1t5d0s0
#metaroot
d11 1 1 c0t0d0s0
d12 1 1 c1t4d0s0
d13 -m d11
#/cms
d19 2 1 /dev/rdsk/c0t0d0s3 1 /dev/rdsk/c0t1d0s1
d20 2 1 /dev/rdsk/c1t4d0s3 1 /dev/rdsk/c1t5d0s1
d21 -m d19
____________________________________________________ 
						

							  Troubleshooting  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Solstice DiskSuite problems81
4
Troubleshooting4
This chapter contains procedures for troubleshooting problems 
you may experience while installing or maintaining systems 
that use disk mirroring.
Solstice DiskSuite problems4
The root filesystem 
runs out of space 
during a system 
install
4
If you run out of space on the /(root) file system—while trying to install 
Solaris patches, for example—you probably installed your /cms swap file 
on 
root. Do the following to recover:
1. Unmount 
/cms:umount /cms
2. Enter the command:ls -l /cms
If the output of that command lists a swap file, do the following:
a. Enter:mount /cms
Remounts /cms.
b. Enter:df -k
Verifies that space is available.
Excessively long 
resync
4
When using metadetach and metattach to detach and then reattach a 
submirror, it may take a long time—sometimes hours—for resync to 
complete. You cannot do any maintenance (adding or replacing disks and 
so on) while the resync is in progress. That behavior is normal. Try again 
later. 
						

							  Troubleshooting  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
The system fails to recognize all disk drives82
The system fails to recognize all disk drives4
If the system should fail to recognize all disk drives, do the following:
1. Reboot the system by entering the following command:
The system reboots and displays the ok
 prompt:
2. Turn off the system unit.
3. Turn off all external devices starting with the system monitor and 
working toward the device farthest from the system unit.
4. Check all disk drive connections to make sure they are secure.
5. Turn on the power to the system components in the opposite order 
in which you powered them off. Power on the SCSI devices first, 
beginning with the device at the end of the chain and working your 
way toward the system unit. Then power on other devices, again 
working your way toward the system unit. 
Finally, power on the system unit itself and the system monitor. 
When you power on the system unit, the system begins to boot. 
Interrupt the boot by pressing Stop-A.
 The system displays the ok
 
prompt.
6. To verify that the system sees all the disk devices, including any new 
drives, enter one of the following commands:
If you receive a “This command may hang the system
....” 
message, respond with the following steps:
a. Enter:n
 (to prevent the probe from continuing).
b. Enter:setenv auto-boot? false
(to prevent a reboot when you do the next command)
c. Enter:reset-all
(it may take a minute to complete).
Now repeat the probe-scsi-all
 or probe-fcal-all
.
The program responds to the probe as shown below. The specific 
response, of course, differs depending on the platform.# /usr/sbin/shutdown -y -g0 -i0
ok probe-scsi-all
ok probe-fcal-allEnterprise 3000
Enterprise 3500 
						

							  Troubleshooting  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
The system fails to recognize all disk drives83
7.IMPORTANT! If you had to do a reset-all
 command, reset 
autoboot? to true by entering the following command:
Failure to reset 
autoboot? before the next reboot will cause the 
reboot to stop at the boot prompt instead of proceeding through the 
normal boot-up.
8. When you have verified that the system is recognizing all its disk 
drives, reboot the system with a boot -r
 command and log in as 
the root user. For example:/sbus@3,0/SUNW,fas@3,8800000
Target 0
Unit 0 Disk SEAGATE ST34371W 
SUN4.2G74629737K6748
Copyright(c) 1997 Seagate
All rights reserved
.
.
.
Target 6
Unit 0 Disk Removable Read Only device 
TOSHIBA XMS5701TASUN12XCD099704/09/97
ok
ok setenv auto-boot? true
ok> boot -r
.
.
.
console login: root
Password:
!WARNING: 
						

							  Troubleshooting  CentreVu CMS Release 3 Version 8 Disk-Mirrored Systems
Fixing “needs maintenance” messages84
Fixing “needs maintenance” messages4
On rare occasions mirrors may get severely out of sync. An extreme 
example is when a disk is removed from a powered system without first 
being detached with a metadetach
 command. The situation is indicated 
by the appearance of a 
Needs maintenance message in a State line of a 
metastat
 response. For example:
In such situations the metareplace
 command can force the recovery of 
the failed slice. The 
Invoke line of the metastat
 response leads the way. 
Given the response above, for example, you would enter the following:# metastat
d13: Mirror
    Submirror 0: d11
      State: Needs maintenance
    Submirror 1: d12
      State: 
    Pass: 1
    Read option: roundrobin (default)
    Write option: parallel (default)
    Size: 2209680 blocks
d11: Submirror of d13
    State: Needs maintenance
Invoke: metareplace d13 c0t0d0s0
 Size: 2209680 blocks
    Stripe 0:
Device Start Block Dbase State Hot Spare
c0t0d0s0 0  No  Okay
...
#
# metareplace -e d13 c0t0d0s0
# metastat
d13: Mirror
Submirror 0: d11
State: Resyncing    
Submirror 1: d12
State: Needs maintenance 
...
#