Document fins/I0421-1
FIN #: I0421-1
SYNOPSIS: abbrev. Recovering from a root filesystem disk failure-on an internal
FCAL boot disk in a E3500 Server
DATE: Sep/11/98
KEYWORDS: abbrev. Recovering from a root filesystem disk failure-on an internal
FCAL boot disk in a E3500 Server
---------------------------------------------------------------------
- Sun Proprietary/Confidential: Internal Use Only -
---------------------------------------------------------------------
FIELD INFORMATION NOTICE
(For Authorized Distribution by SunService)
SYNOPSIS: Recovering from a root filesystem disk failure or corruption
on an internal FCAL boot disk in an E3500 Server or a FCAL
boot disk in a A5000 which is not mirrored with SEVM or SDS.
TOP FIN/FCO REPORT: Yes
PRODUCT_REFERENCE: FCAL disk recovery
PRODUCT CATEGORY: Storage / / Disk / A5000
PRODUCTS AFFECTED:
Mktg ID Platform Model Description Serial Number
------- -------- ----- ----------- -------------
Systems Affected
------- --------
- E3500 - Ultra Enterprise 3500 -
X-options Affected
--------- --------
SG-XARY010A-45G - - 45.5 GB Sun Enterprise Network Array -
SG-XARY010A-127G - - 127.4 GB Sun Enterprise Network Array -
SG-XARY011A-127G - - 127.4 GB Sun Enterprise Network Array -
SG-XARY012A-254G - - 254.8 GB Sun Enterprise Network Array -
SG-XARY012A-509G - - 509.6 GB Sun Enterprise Network Array -
PART NUMBERS AFFECTED:
Part Number Description Model
----------- ----------- -----
540-3852-0X ASSY,9GB FC-AL W/SPUD BRKT -
540-3249-0X ASSY,9GB FC-AL W/SPUD BRKT -
(SCSI Devices)
Type Vendor Model Serial Number(Min) Serial Number(Max) Firmware
---- ------ ------- ------------------ ------------------ --------
9GB Seagate ST19171FC - - -
REFERENCES:
BugId: 4161768
PROBLEM DESCRIPTION:
There is currently no documented procedure on how to recover from a
failed or corrupted root system disk which has been built on a internal
FCAL disk on an E3500 or a FCAL disk in an A5000 (Photon) enclosure.
To resolve this situation the following recovery procedure is being
provided until such time as one is provided in the customer / service
documentation.
IMPLEMENTATION:
---
| | MANDATORY (Fully Pro-Active)
---
---
| | CONTROLLED PRO-ACTIVE (per Sun Geo Plan)
---
---
| X | REACTIVE (As Required)
---
CORRECTIVE ACTION:
This procedure should be used to recover from a situation where you
have a faulty internal E3500 FCAL boot disk or a boot disk which is in
a A5000 array where the boot disk is NOT mirrored or under SEVM or SDS
control.
In the following recovery procedure a Solaris 2.6 HW 5/98 CDrom media
was used as the boot device and Solaris recovery image.
It has been reasonably assumed the system is down and at the ok prompt
as the root disk is faulty and waiting to be recovered.
1. Power off the E3500 or A5000 Array with the faulty boot disk
2. Using appropriate anti-static precautions remove the faulty boot
disk and insert a good replacement disk (make a note of the WWN of
the of the new replacement disk as found on the printed disk label,
you will need this later in the procedure.) Return AC Power to the
E3500 or A5000 array.
3. After inserting the appropriate boot media in to the CDrom drive
attached to the host system from the ok: prompt type boot cdrom -sw
4. When the system reaches the # prompt type the following command:
# format
(from the list provided select the replacement boot drive, it
should be easily identified by its WWN. Then partition it as per the
removed faulty drive. System admins should of course keep a print out
of the format table or vtoc of the root disk drive.)
(Also you will need to make a note of the full path to the device as
displayed by format as you will need it later to change the OBP boot
device path)
The path noted should be something like this:
/sbus@2,0/SUNW,socal@2,0/sf@0,0/ssd@w21000020370793ce,0
5. Once partitioned and labeled via format each slice should have newfs
run on it.
# newfs /dev/rdsk/cXtXdXs0 (root slice)
# fsck /dev/rdsk/cXtXdXs0
# newfs /dev/rdsk/cXtXdXs6 (/usr slice)
# fsck /dev/rdsk/cXtXdXs6
etc. until all required slices are newfs'd and fsck'd
other partitions required may be /opt /var /export/home etc
6. Now mount what is to be the root partition onto a temporary mount point
e.g. /a
# mount dev/dsk/cXtXdXs0 /a
After inserting the backup tape restore your root file system slice backup
from the tape media.
# cd /a
# ufsrestore rvf /dev/rmt/X
# rm /a/restoresymtable
7. Now mount what is to be the /usr partition onto your temporary mount point
/a/usr
# mount dev/dsk/cXtXdXs6 /a/usr
# cd /a/usr
# ufsrestore rvf /dev/rmt/X (you may have to position to a second logical
tape using the mt command or put in another
tape to recover each different filesystem)
# rm /a/usr/restoresymtable
8. Continue this mount and restore operation until all the required
root disk partitions are recovered.
9. Once all the restores are complete the following commands should
be run. These commands will build in the WWN of the replacement
disk into the relevant paths so that the system will be able
to boot the operating system from the new disk.
# drvconfig -r /a/devices -p /a/etc/path_to_inst
# disks -r /a
# devlinks -r /a
* The next two commands are currently required as a result of bugID 4161768
as the "drvconfig -r -p" command currently does not correctly build
in the
FCAL device paths in /devices, once the bug is resolved these steps may be
omitted.
# cd /devices
# find . -print | cpio -pduvm /a/devices
10. Install the bootblock to the replacement disk
# installboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/cXtXdxs0
11. Umount the filesystems /a/usr .. /a/xx /a etc
12. fsck the filesystems to set the inode table correct after the
restore
# fsck /dev/rdsk/cXtXdXs0
# fsck /dev/rdsk/cXtXdXs6
etc.
etc.
until all the newly restored filesystems are clean.
13. Create the correct boot device using the luxadm command
# luxadm -v set_boot_dev -y /dev/dsk/cXtXdXs0
Current boot-device =
/sbus@2,0/SUNW,socal@2,0/sf@0,0/ssd@w21000020370e1814,0:a
New boot-device = /sbus@2,0/SUNW,socal@2,0/sf@0,0/ssd@w21000020370793ce,0:a
14. Shutdown the system cleanly. Your system should now boot from the
replacement disk. At the ok prompt type:
ok boot
COMMENTS:
------------------------------------------------------------------------
Implementation Footnote:
________________________
i) In case of MANDATORY FINs, Enterprise Services will attempt to contact
all affected customers to recommend implementation of the FIN.
ii) For CONTROLLED PROACTIVE FINs, Enterprise Services mission critical sup-
port teams will recommend implementation of the FIN (to their respective
accounts), at the convenience of the customer.
iii) For REACTIVE FINs, Enterprise Services will implement the FIN as the need
arises.
--------------------------------------------------------------------------
All released FINs and FCOs can be accessed using your favorite network
browser as follows:
SunWeb Access:
______________
* Access the top level URL of http://cte.corp/FIN_FCO/
* From there, select the appropriate link to query or browse the FIN and
FCO Homepage collections.
Supporting documents for FIN/FCOs can be found on Edist. Edist can be
accessed internally at the following URL: http://edist.corp/.
* From there, follow the hyperlink path of "SunService Documentation"
and
click on "FIN & FCO attachments", then choose the appropriate
folder, FIN or
FCO. This will display supporting directories/files for FINs or FCOs.
Internet Access:
_______________
* Access the top level URL of https://infoserver.Sun.COM
--------------------------------------------------------------------------
General:
________
Send questions or comments to finfco-manager@cte.Corp
---------------------------------------------------------------------------
Copyright (c) 1997-2003 Sun Microsystems, Inc.