- Community Home
- >
- Servers and Operating Systems
- >
- Operating Systems
- >
- Operating System - HP-UX
- >
- Re: BAD BLOCKS
Categories
Company
Local Language
Forums
Discussions
Forums
- Data Protection and Retention
- Entry Storage Systems
- Legacy
- Midrange and Enterprise Storage
- Storage Networking
- HPE Nimble Storage
Discussions
Forums
Discussions
Discussions
Discussions
Forums
Discussions
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
- BladeSystem Infrastructure and Application Solutions
- Appliance Servers
- Alpha Servers
- BackOffice Products
- Internet Products
- HPE 9000 and HPE e3000 Servers
- Networking
- Netservers
- Secure OS Software for Linux
- Server Management (Insight Manager 7)
- Windows Server 2003
- Operating System - Tru64 Unix
- ProLiant Deployment and Provisioning
- Linux-Based Community / Regional
- Microsoft System Center Integration
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Community
Resources
Forums
Blogs
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 12:56 AM
08-22-2006 12:56 AM
BAD BLOCKS
Environment 11.11
Database folks are complaining of BAD BLOCKS and they refreshed the database. from the OS side i recommended umounting and doing full fsck on the filesystem in question. Now the customer wants to know the root cause of the issue. lvdisplay, syslog dmesg are all clean. Anything else we can check to provide the root cause for the BAD BLOCKS that oracle experienced?
Thanks
Brian
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:03 AM
08-22-2006 01:03 AM
Re: BAD BLOCKS
anything showing with:
diskinfo -v /dev/rdsk/cxtxdx
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:13 AM
08-22-2006 01:13 AM
Re: BAD BLOCKS
If you looking for Local then you can check using dd command or ask hp and they can give a binary program which checks bad blocks and also it may be fixed.
dd if=/dev/rdsk/cXtYdZ of=/dev/null bs=4096k
If it is SAN disk then talk to the Vendor.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:27 AM
08-22-2006 01:27 AM
Re: BAD BLOCKS
# echo 2400?20X|adb /dev/dsk/c2t0d0
2400: 44454645 43543031 0 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
#
dd on a failed disk will hang for ever!!
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:47 AM
08-22-2006 01:47 AM
Re: BAD BLOCKS
Oracle will probably not tolerate OS bad block relocation.
This is an oracle issue and the database may need to be restored from backup.
If you have archive logging turned on there should not be much data loss.
SEP
Owner of ISN Corporation
http://isnamerica.com
http://hpuxconsulting.com
Sponsor: http://hpux.ws
Twitter: http://twitter.com/hpuxlinux
Founder http://newdatacloud.com
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:57 AM
08-22-2006 01:57 AM
Re: BAD BLOCKS
Is this connected to some type of disk array such as an EMC Symmetrix? Most modern arrays will handle bad block relocation at the hardware level. If you have BB relocation set on an LV and the array detects and moves the block there will be problems. Basically LVM moves the block as does the array so now you have a lost block.
The last time I ran into this the DBAs had to do some recovery. For any LV's built on EMC Symmetrix LUNs you should have BB relocation set to NONE (lvchange -r N /dev/
I'm sure other arrays (Hitachi, etc.) are similar.
David
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 01:58 AM
08-22-2006 01:58 AM
Re: BAD BLOCKS
If the condition I just talked about happens you will not find any bad blocks on your LV because the array already handled it. Your data block has been moved however so that block would be empty.
David
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 02:22 AM
08-22-2006 02:22 AM
Re: BAD BLOCKS
Thanks
Brian
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 07:17 AM
08-22-2006 07:17 AM
Re: BAD BLOCKS
For Symmetrix it needs to be set to NONE (lvchange -r N /dev/
From the lvchange man page:
"-r N Disable bad block relocation and the Bad Block Directory. Upon a media failure, LVM will NOT attempt to relocate the bad block. In addition it will NOT enter the block in the Bad Block Directory. LVM will have no record of the block being bad, and will attempt to access it on future I/O requests."
This way the array can handle any bad blocks and present only good blocks to the host.
David
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 07:27 AM
08-22-2006 07:27 AM
Re: BAD BLOCKS
You can have bad blocks at two different levels, at the file system/hard drive level, and also at the Oracle level. Are the Oracle folks complaining of ORACLE BAD BLOCKS? If so, they can verify the files and where the errors are with the "dbv" tool (from Oracle). Syntax is :
dbv BLOCKSIZE=[YOUR SIZE] FILE=[your file]
e.g.
dbv BLOCKSIZE=8192 FILE=user01.dbf
WARNING: THE ABOVE COMMAND IS *DBA FUNCTION* - not for sysadmin consumption!!!
If this shows an error in a file, you can quickly determine if this is a UNIX issue by running a dd from the file to /dev/null:
dd if=user01.dbf of=/dev/null bs=8k
If there is an error, it should give you an error in the same block as the Oracle error. If it is clean from the dd command, it is very likely that this is an ORACLE ONLY problem, and really much to do with you (other than helping to restore if necessary).
Have the dba's clarify how the BAD BLOCK is manifesting itself, and check it out a bit yourself.
John
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 08:33 PM
08-22-2006 08:33 PM
Re: BAD BLOCKS
I am not sure where your disks are connected to, either SAN or local internal disk?
If it is a SAN disk it is recommended to turn off the bad block allocation on the logical system level. You SAN will manage bad block allocation by default.
There are tools available (I am not sure who will supply either HP or SAN vendor) to check bad block reallocation on the logical volme level.
To check bad blocks on the SAN, you have to contact your storage vendor to check it.
Short and sweet, turn off bad block reallocation on the logical volume level if connected to SAN.
Regards
Sathish
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-22-2006 08:39 PM
08-22-2006 08:39 PM
Re: BAD BLOCKS
maybe offtopic, but here is my ksh script for checking bad block setting:
#!/usr/bin/ksh
# writed by Marek Mahut,
# This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY !
#
echo "\nThis script show bad block settings for each LV: "
ktosom=`/usr/bin/whoami`
if [[ $ktosom = "root" ]]; then
getvgs=`ls -l /dev/*/group | awk '{print $10}' | sed 's/group//g'`
for var in $getvgs ; do
num=`ls -l $var | sort -i | awk '{print $10}' | grep "^lvol" | wc -l`
if [ $num -eq 0 ]; then
bla="blabla"
#echo "\n==>> VG $var has not logical volumes."
else
echo "\n==> Logical volumes for VG $var: "
fi
volu=`ls -l $var | awk '{print $10}' | grep "^lvol" | sort -n`
for vat in $volu ; do
getsetting=`/usr/sbin/lvdisplay -v $var$vat | grep -i "bad block" | awk '{print $3}'`
echo "-> Bad block setting for $var$vat is set to \033[0;1;40;1m$getsetting\033[0m";
done
done
else
echo "\n\033[1;37m\033[41m ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! \033[m"
echo "\033[1;37m\033[41m ! ! ! You are $ktosom, you don't have permissions to run this script ! ! \033[m"
echo "\033[1;37m\033[41m ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! \033[m\n"
fi
echo "\nAnd have a nice day! :)"
all suggestions welcome :)
Marek.