Flash archive from a T2000 to V240

Hi,
I am attempting to install a flash archive created on a sun T2000 and install it on a Sun V240. Its reporting errors that say the flash archive is not valid. I have tried many times. I don't quite understand this sun4u vs sun4v stuff. Is the T2000 sun4v and the v240 sun4u? and if so, is it possible to flash archive between them?
Any help on this would be great!
Thanks,
Darren

Sorry. Thank you for answering my question.
I have formated my flash drive to a Fat 32, and there is no difference.
I save file on my flash drive and when i try to re-open it it says : File Error: Couldn't open the file. It may be corrupt or a file format that Preview doesn't recognize. Sometimes the same file will save and open perfectly, and other times the error message comes up.
Thanks.

Similar Messages

Solaris 8 container install problem using Ultra 60 flash archive

Hi
I have an M4000 configured with a Solaris 8 container.
I am using a flash archive from an Ultra 60.
Below is the problem when I install.
# zoneadm -z f714clone install -u -a /var/flar/fid714_20090515.flar
Log File: /var/tmp/f714clone.install.26921.log
Source: /var/flar/fid714_20090515.flar
Installing: This may take several minutes...
Postprocessing: This may take several minutes...
WARNING: zone did not finish booting.
Last entry in the logfile is...
[Fri May 15 12:21:22 BST 2009] Postprocess: Performing zone sys-unconfig
It just sits at the WARNING doing nothing.
I connected to the container console using "zlogin -C f714clone" but this just gives me a blank screen with no output and unable to do anything.
I used a different archive from a V120 and had no problems.
Is there something I need to configure for a Ultra 60.
Thanks in advance.
Steve.

I got this to work. It seems, for whatever reason, that it doesn't like the '-u' flag for an Ultra 60. I installed using '-p' instead and it was fine, it even came up with the new IP I had assinged.
zoneadm -z QeSun01b install -p -a qesun01.flar

Errors when attempting to Jumpstart install a flash archive: Solaris10_u8

I pretty familiar with Jumpstart. I am attempting to use the new feature in Solaris10_u8 (10/2009) where you can do a flash archive install for zfs root systems.
I first built a machine the way I wanted it, and applied all patches. Then I ran flarcreate to get my new image. Copied the new image to an HTTP server.
Then I changed my rules/profile file to this time install the flash archive. Here is what my new profile looks like:
install_type flash_install
archive_location http://nas02.clark-communications.com/solaris/solaris10_u8/flash/sol101009flar
partitioning explicit
pool rpool auto auto auto mirror c0d0s0 c1d0s0
The jumpstart install begins fine.
At some point in the install, I see this:
Extracted 1632.03 MB ( 90% of 1797.45 MB archive)
cannot receive new filesystem stream: invalid stream (checksum mismatch)
Extracted 1633.03 MB ( 90% of 1797.45 MB archive)
The extraction continues. Later on, I see this:
ERROR: Could not stop the extraction
ERROR: Could not extract Flash archive
ERROR: Flash installation failed
Solaris installation program exited.
And the installation fails/stops.
What am I doing wrong, how do I fix this?

I've tried this both ways- using Solairs 10 update 6 at least, installing a fash archive created on a zfs-root system to a ufs filesystem AND ainstalling a flash archive from a ufs based root onto a zfs filesystem both failed with odd errors similar to what you experienced. I have not tried this again with either update 7 or update 8, but I believe you will encounter the same problem.
As far as I've seen so far- if the flash was created on a zfs root- you have to extract it to a zfs root filesystem, and if the original was ufs, you can only extract it to a ufs filesystem.

How to take flash archive zoned system and how to deployee from jumpstart s

Hi All
How to take flash archive zoned system and that flash archive how to deployee from Jumpstart server
please help me in this,it is very urgent for me
Thanks
Murali

when you've created the flar image just setup the jumpstart server accordingly
1) create a profile for a flash installation
install_type flash_install
archive_location nfs js1:/data/server.flar
partitioning explicit
filesys c0t0d0s0 2000 /
filesys c0t0d0s3 6000 /usr
filesys c0t0d0s4 2000 /var
filesys c0t0d0s5 2000 /export/home
filesys c0t0d0s6 520000 /export/zones
2) create the jumpstart client just as you would normally
3) jumpstart the server
As far as i know this should work

Flash archive install from local_device

I created a flash archive image from a server, now i want to install a new server using this file from a local scsi disk , but i have no idea on how to do this. I found info about network install, but i can't use network for this install. Any help will be wellcome,
thanks in advance.
Andres.

Hi,
in first place thanks for your help. I booted from CD, and gave the location of my flash image, then started the installation, but, about 60 % of execution, i got an error message like this:
fopen: No such file or directory
ERROR: Unable to unconfigure the extracted system
Unable to unconfigure the extracted system
And the system hanged. After that i have to make Stop + A to recover control on the system.
so, do you have any idea about this message?

Flash archive problem

I upgraded my T2000 system from Solaris 10 u1 to Solaris 10 u3. Now when I create a flash archive. The "content_architectures" field in flar info contains "sun4c,sun4d,sun4m,sun4u,sun4us,sun4s", and I cannot jumpstart another T2000 with it. any ideas?

Just encountered this same problem with an X64 build using a flash archive of Solaris 10/09 and using XVM 2.5.0.1186 :
Installing boot information
cannot mount '/a/rpool': directory is not empty
bootadm: mount of ZFS pool rpool failed
bootadm: cannot mount pool dataset for pool: rpoolWhen I examine the canmount zfs options just after it fails I see :
# zfs get canmount
NAME                PROPERTY VALUE     SOURCE
rpool               canmount noauto    local
rpool/ROOT          canmount on        default
rpool/ROOT/zfsroot canmount noauto    local
rpool/dump          canmount -         -
rpool/swap          canmount -         -Should I be setting canmount=on for just 'rpool' then ?

Flash Archive restored files

Hi,
We are moving servers (T2000' / Solaris10) to a new Datacentre and using swing kit to reduce the downtime on the production databases.
We plan to take a Flash Archive of the OS and restore the swing kit (exactly the same config) from the Flash Archive, we also plan to take ufsdumps.
Would anyone know the exact files which are not restored as part of the Flash Archive please?? I know Flash Archive runs a sys-unconfig so hosts, defaultrouter, defaultdomain and the Network will not be restored but are there any other files which will not be restored???
Thanks in advance.

LDAP may or may not work. I have three T2000s that I need to update and I've flar'ed them and restored them to a matching T2000 just so there won't be any surprises when I do the real boxes. Sometimes LDAP works, sometimes it doesn't.

FLASH ARCHIVE INSTALATION FAIL

Hi guys,
I am just trying to install the following Flash Archive:
bash-2.05# flar -i Cms
archive_id=792943ac9a79c51395a2244413ca7bc4
files_archived_method=cpio
creation_date=20050729163902
creation_master=cmssun1
content_name=Solaris9 + Operations 8.12
creation_node=cmssun1
creation_hardware_class=sun4u
creation_platform=SUNW,Sun-Fire-V240
creation_processor=sparc
creation_release=5.9
creation_os_name=SunOS
creation_os_version=Generic_118558-09
files_compressed_method=compress
content_author=RH & RJH
content_architectures=sun4u
type=FULL
bash-2.05# ls -la Cms
-rw-r--r-- 1 root other 3458497988 Apr 27 11:32 Cms
bash-2.05# file Cms
Cms: ascii text
After booting from CD, I select Flash Archive Installation -> From Local device. I select the correct device, and when I point to the file /Cms, the program says: "could not find archive".
This Flash Archive was created on a SUNW,Sun-Fire-V240. The system I am trying to install is a SUNW,Ultra-60.
I appreciate any help.
Thanks in advance,
Ricardo

When you boot off a CD, the hard drive root file system normally is mounted under /a, assuming the image is on the system that you want to install the image on..

Flash archive to zone?

Is there a way to take a flar from a solaris 10 physical machine and install it into a zone of another machine (also solaris 10)? I see the migration software for solaris 8 to a container, but not solaris 10 to container.

redhat_123 wrote:
Even if we have a whole root zone, extracting a flash archive wouldnt help. We really need to wait for Sun to provide a native P2V for Sol10. There would be work arounds, but not going to be supported by Sun.I'm not sure there's value in P2V for Solaris 10. IMO that would be handled just by building the zone and migrating what you've got to the zone.
I'm sure I'm missing something here... the question is what.
Best,

Flash archive on solaris 8/9

Hello,
I have a nfs server containing flar image but this box is not a boot server.
I need to boot from CDROM and recover OS from a flash archive server (NFS). I have all the files in place too.
what command do i need to run from the ok prompt of the corrupt server so that it boot initially for CD Rom and installs from the flash archive?
ok> boot cdrom - install net=nfs_servername:/path_to_archive
or
ok> boot cdrom - install nfs=nfs_servername:/path_to_archive
please help i have have been searching a lot bot with no luck please help.

I have seen
boot cdrom - install tape=/dev/rmt/0n
i have seen this in man of boot
ok boot cdrom -F wanboot - install
and saw this on some fourm
boot cdrom -o prompt -F wanboot -V - install
but not sure how do i use them though.
and also dont know where i can specify the nfs server and archive path at boot prompt.
Any ideas?

Solaris 10 physical to Solaris 10 zones via flash archive

Hi,
I need to migrate E25K domains (all Solaris 10) to zones on an M9000. The solution architect has suggested zones on the M9000 made from flash archives of the source E25K domains for speed & simplicity (read : avoid a rebuild of the application environment and all that entails). But as best I can tell, you can only use flash archives to provision zones as part of the branded zones architecture and it does not include SOlaris 10.
As best I can tell, this is not possible - at least not in a supported way; I have seen some blogs that have hacks for making a zone from a solaris 10 flash archive but I can't do that sort of thing here.
Is my view correct? If it is, is there any timeframe for Solaris 10 P2V?
I am aware that OpenSolaris now supports Solaris (TM) as a branded zone but that will not help me here.
TIA
Steve

You log an RFE the same way you log a bug with Sun/Oracle, by calling the Support Center. For that, you would need a support contract or server serial number. If you have a relationship with a Sun/Oracle sales team, you may also be able to speak with them directly and ask if they could open the call on your behalf. The call that you open is investigated and may be determined by Sun/Oracle to be an RFE. I think I mentioned earlier that the Solari10 P2V may already have been requested by various customers, but it is always good to log the call and have yourself added as an interested party.

Solaris 9 U7 (2004/09) MultiLayered Flash Archive issue + solution ?

In Solaris 9 Update 7 (2004/09) the multilayered flash archive installation feature is broken.
I think that this is a known issue:
4945516 Failure of multi-layered archive with Solaris 9 4/03 on Postdeployment process
I manage to flash my archives by changing the following statement found at line 24 of the file
../Solaris_9/Tools/Boot/usr/lib/flash/postdeployment_processing
from
if [ ! -L $FLASH_ROOT/etc/rc2.d/S98flash_reboot ] ; then
to
if [ -L $FLASH_ROOT/etc/rc2.d/S98flash_reboot ] ; then
This issue is addressed in the patch 113434-22
113434-22/SUNWinst/reloc/usr/lib/flash/postdeployment_processing
But the if statement is not the good one.
Is this correct ?
Coco D.

Ideas.. Hmm, none which seems that correct, but you could try some things.
If it gets the wrong IP that could explain why it fails to copy the sysidcfg file.
First you could try and do a snoop on the ethernet address;
snoop ether 0:3:ba:14:c6:cd
(you could also try the -v flag to increase the verbosity).
The things you should look for are arp/rarp requests, the jumpstart client will use arp/rarp to determine its IP address, snoop will show you which server that responds, and what address it gets, futher more its a good idea to verify that the response to the bootparams requests comes from the correct server.
Of course there might be other oddnesses as well.
Was the data you provided extracts? If it was you should check the /etc/ethers and /etc/bootparams for duplicates entries, so there are no other occourances of the clients ehternet address in /etc/ethers, and no bogus entries starting with * or the same hostname in /etc/bootparams.
If you added the client manually you could always try and use the Tools/rm_install_client and Tools/add_install_client scripts to add it again, these script sometimes detect problems with the configuration.
Lastly you didn't show us your /etc/nsswitch.conf file, but i assume that it has "files" first for the ethers, hosts and bootparams entries?
Good luck
//Magnus

Are nfs mounts allowed in vfstab after a flash archive restore?

I am restoring a server using a flash archive. The vfstab on the original server contains an entry for an nfs mount. It looks like during the push of the archive to the new server, the vfstab gets saved to vfstab.orig and the nfs mount is removed from vfstab. So I added a finish script to add the nfs mount back to vfstab and now the server gets hung during S73nfs.client during reboot.
Are nfs mounts not allowed in vfstab after a flash archive restore?

One option is to secure the OBP with a password.
At the OK prompt, set the security-mode to "command" and then set a password. Careful not to loose this password. With this set the only option is to "boot" without supplying the firmware password.
Commands:
# halt
ok setenv security-mode = command
ok setent security-password <enterpasswd>
ok reset-all

Restoring complete flash archive

Recently I made a flash archive of my system solaris 8 to tape using the command:
FLARCREATE -N SUN-EC1.FLAR -C -S -R / -T /DEV/RMT/0
I have 2 disks of 32 GB and I will replace them for 2 of 72 GB.
Which command should I use to restore my system completely after the change of the disks?
The image was taken from a live system ; Can I have some problem when restoring the copy?

To restore the backup made with flarcreate you need to boot your upgraded server with the Installation disk CD1
I used the form on the <OK> prompt:
boot cdrom - w
Then the installation should start whare you have to define slices and network parameters and then you should be offered to install OS from standard location (cd) or flash archive. There you need to select flash_archive (I think it is Cntr-4 of F4) which should guide you to select source (local tape) and you just have to confirm and sit down and watch.
Good luck!
Zika from RSA

Script remove archives from standby after applied

Hello,
I am looking for a script to remove archivelog files from the standby flash recovery area after beeing applied.
My database is 11.2.0.3.3 on Linux Redhat 5.
Can someone kindly proovide me an example?
Thank you

RMAN does that for you :
RMAN>CONFIGURE ARCHIVELOG DELETION POLICY TO APPLIED ON STANDBY;
Then set the date to what you need and run as often as you need.
http://docs.oracle.com/cd/E18283_01/server.112/e17022/rman.htm#BAJHHAEB
#!/bin/bash
# Name: db_rman_arch_standby.sh
# Purpose: Delete archive from standby
# Usage : db_rman_arch_standby <DBNAME>
if [ "$1" ]
then DBNAME=$1
else
echo "basename $0 : Syntax error : use . db_rman_full <DBNAME> "
exit 1
fi
. /u01/app/oracle/dba_tool/env/${DBNAME}.env
echo ${DBNAME}
MAILHEADER="Archive_cleanup_on_STANDBY_${DBNAME}"
echo "Starting RMAN..."
$ORACLE_HOME/bin/rman target / catalog rmancat/rmancat@rcatalog << EOF
delete noprompt ARCHIVELOG UNTIL TIME 'SYSDATE-1';
exit
EOF
echo `date`
echo
echo 'End of archive cleanup on STANDBY'
mailx -s ${MAILHEADER} $MAILTO < /tmp/rmandbarchstandby.out
. /u01/app/oracle/dba_tool/bin/rmemptyfolders.sh ${DBNAME}
# End of ScriptUses an ENV file ( standby.env ) for each database. Use Linux ENV to compare to your settings
ORACLE_BASE=/u01/app/oracle
ULIMIT=unlimited
ORACLE_SID=STANDBY
ORACLE_HOME=$ORACLE_BASE/product/11.2.0.2
ORA_NLS33=$ORACLE_HOME/ocommon/nls/admin/data
LD_LIBRARY_PATH=$ORACLE_HOME/lib:/lib:/usr/lib
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
LIBPATH=$LD_LIBRARY_PATH:/usr/lib
TNS_ADMIN=$ORACLE_HOME/network/admin
PATH=$ORACLE_HOME/bin:$ORACLE_BASE/dba_tool/bin:/bin:/usr/bin:/usr/ccs/bin:/etc:/usr/sbin:/usr/ucb:$HOME/bin:/usr/bin/X11:/sbin:/usr/lbin:/GNU/bin/make:/u01/app/oracle/dba_tool/bin:/home/oracle/utils/SCRIPTS:/usr/local/bin:.
export TERM=vt100
export ORACLE_BASE ORACLE_SID ORACLE_TERM ULIMIT
export ORACLE_HOME
export LIBPATH LD_LIBRARY_PATH ORA_NLS33
export TNS_ADMIN
export PATHBest Regards
mseberg
Edited by: mseberg on Oct 25, 2012 4:58 AM

Flash archive from a T2000 to V240

Similar Messages

Maybe you are looking for