Low flash error after upgrade 6.1.3 to 6.4.4

[rdehaan]

Dear all,

After uprgrading from 6.1.3.693.R01 GA, October 17, 2006 to Alcatel-Lucent OS6850-24 6.4.4.402.R01 Service Release, October 10, 2011 i have a low memory problem preventing me from certify the switch. The ls command is telling me that there is 3 MB free. (with 6.1.3.693.R01 GA there is 26 MB free)
Is there a way to relase or add some memory?

-> write memory
File /flash/working/boot.cfg replaced.
This file may be overwritten if "takeover" is executed before "certify"
-> copy working certified
ERROR : Synchronization Commands will not be accepted due to low flash . Remove unwanted files
CERTIFY process failed
ERROR: Attempt to copy working to cert w/flash synchr failed
Not enough flash memory!

ls output
-> ls -r

Listing Directory /flash:

drw 512 Jan 29 08:11 certified/
-rw 338 Jan 29 08:22 boot.params
drw 512 Feb 7 11:57 working/
-rw 11 Jul 24 2006 boot.slot.cfg
-rw 64000 Feb 11 07:32 swlog1.log
-rw 64000 Feb 14 12:16 swlog2.log
drw 512 Jan 28 21:49 network/
drw 512 Jan 29 08:32 switch/
-rw 20 Jan 28 21:48 installed

Listing Directory /flash/certified:

drw 512 Jan 29 08:11 ./
drw 512 Feb 7 10:00 ../
-rw 1894790 Jan 28 21:42 K2os.img
-rw 5712300 Jan 28 21:45 Keni.img
-rw 18324890 Jan 28 21:44 Kbase.img
-rw 632708 Jan 28 21:45 Ksecu.img
-rw 785 Oct 19 2006 software2.lsm
-rw 7188 Jan 28 21:37 boot.cfg
-rw 2864365 Jan 28 21:42 Kadvrout.img
-rw 3437 Jan 28 21:44 Kencrypt.img

Listing Directory /flash/working:

drw 512 Feb 7 11:57 ./
drw 512 Feb 7 10:00 ../
-rw 1894790 Jan 28 21:42 K2os.img
-rw 7388 Feb 7 11:57 boot.cfg
-rw 5712300 Jan 28 21:45 Keni.img
-rw 18324890 Jan 28 21:44 Kbase.img
-rw 632708 Jan 28 21:45 Ksecu.img
-rw 785 Oct 19 2006 software2.lsm
-rw 2864365 Jan 28 21:42 Kadvrout.img
-rw 3437 Jan 28 21:44 Kencrypt.img

Listing Directory /flash/network:

drw 512 Jan 28 21:49 ./
drw 512 Feb 7 10:00 ../
-rw 23040 May 27 2009 userTable4
-rw 33 Jun 27 2007 policy.cfg
-rw 404 Jun 27 2007 ssh_host_dsa_key
-rw 358 Jun 27 2007 ssh_host_dsa_key.pub
-rw 91392 Jan 28 21:49 userTable5
-rw 12 Jan 28 21:49 lockoutSetting

Listing Directory /flash/switch:

drw 512 Jan 29 08:32 ./
drw 512 Feb 7 10:00 ../
-rw 105613 Sep 8 2006 certs.pem
-rw 20 Jan 29 08:22 snmp.engine
-rw 0 Sep 8 2006 pre_banner.txt
drw 512 Jan 28 21:48 avlan/
-rw 1732 Jan 28 21:48 label.txt
-rw 944 Sep 8 2006 wv-cert.pem
-rw 981 Sep 8 2006 wv-key.pem
-rw 1732 Jan 4 2007 label.txt.bk
-rw 2642 Jan 29 08:32 dhcpBind.db

Listing Directory /flash/switch/avlan:

drw 512 Jan 28 21:48 ./
drw 512 Jan 29 08:32 ../
-rw 445 Jan 28 21:48 topA.html
-rw 358 Jan 28 21:48 topB.html
-rw 388 Jan 28 21:48 bottomA.html
-rw 384 Jan 28 21:48 bottomB.html
-rw 2444 Jan 28 21:48 myLogo.gif
-rw 8698 Jan 28 21:48 javlan.cab
-rw 13521 Jan 28 21:48 javlan.jar
-rw 1121 Jan 28 21:48 avlanMacInstall.html
-rw 51200 Jan 28 21:48 avlaninstall.tar
-rw 12090 Jan 28 21:48 javlan.jar.sig
-rw 5327 Jan 28 21:48 javlaninstall.sit
-rw 11569 Jan 28 21:48 javlanmacosx.jar
-rw 1722 Jan 28 21:48 reconfigApple.html
-rw 445 Jan 4 2007 topA.html.bk
-rw 358 Jan 4 2007 topB.html.bk
-rw 388 Jan 4 2007 bottomA.html.bk
-rw 384 Jan 4 2007 bottomB.html.bk
-rw 463 Jan 4 2007 myLogo.gif.bk
-rw 8698 Jan 4 2007 javlan.cab.bk
-rw 11995 Jan 4 2007 javlan.jar.bk
-rw 1121 Jan 4 2007 avlanMacInstall.html.bk
-rw 51200 Jan 4 2007 avlaninstall.tar.bk
-rw 12090 Jan 4 2007 javlan.jar.sig.bk
-rw 5327 Jan 4 2007 javlaninstall.sit.bk
-rw 11698 Jan 4 2007 javlanmacosx.jar.bk
-rw 1722 Jan 4 2007 reconfigApple.html.bk

3598848 bytes free

[jinkama-san]

Hi,

is this a stack configuration? If so, check the flash space on all the non-primary stack members individually.
It happened to us before that the flash synchro within the stack would fail because one stack member had an old coredump file in the flash.

[one6f]

Flash is running short
Try to remove old files:
rm /flash/switch/*.bk
rm /flash/switch/avlan/*.bk
According to this post you can try to
rm /flash/network/userTable4

/flash/switch/avlan/avlaninstall.tar ??? I'm not sure.

[rdehaan]

Problem solved

Thank you for the helpful response.
This is what I had to do to solve the low flash problem.

Step 1) Removed the *.bk from /flash/switch/ and /flash/switch/avlan/ from flash on slot 1
Step 2) Removed the files K2diag.img and Krelease.img from /working and /certified on slot 1/5
Step 3) reload working no rollback-timeout

Result :
-> write memory
File /flash/working/boot.cfg replaced.
This file may be overwritten if "takeover" is executed before "certify"
-> copy working certified
Setting CERTIFY Timeout for 800 seconds
from /flash/working to /flash/certified
Copying boot.cfg .................... completed

CERTIFY process completed successfully
Flash Synchronization process started
+++ == CSM == Stack 1 Certify process Completed
+++ == CSM == Stack 3 Certify process Completed
+++ == CSM == Stack 4 Certify process Completed
+++ == CSM == Stack 2 Certify process Completed
+++ == CSM == Stack 5 Certify process Completed
Flash Synchronization process completed successfully

-> freespace /flash
/flash 3706880 bytes free

[bma]

Hello,
I have a similar problem with 4 switch in stack.
I removed the file.bk from flash/switch and flash/switch/avlan
You said: "Removed the files K2diag.img and Krelease.img from /working and /certified on slot 1/5" these files are not in /working and /certified .
The switch booted in certified and when I type this command , there are this message error again!
213# RELOAD WORKING NO ROLLBACK-TIMEOUT
ERROR : Synchronization Commands will not be accepted due to low flash . Remove unwanted files

Command ls -r:
Listing Directory /flash:

drw 1024 May 16 16:07 certified/
-rw 340 May 21 06:53 boot.params
-rw 64000 May 23 17:21 swlog1.log
-rw 64000 May 23 17:03 swlog2.log
drw 1024 Mar 6 18:13 switch/
-rw 11 Dec 27 2008 boot.slot.cfg
drw 1024 Aug 13 2011 network/
-rw 20 Mar 6 18:13 installed
-rw 218 Mar 6 18:14 boot.cfg.1.err
-rw 501 Mar 6 18:04 software.lsm
drw 1024 May 16 15:39 working/

Listing Directory /flash/certified:

drw 1024 May 16 16:07 ./
drw 1024 May 21 06:54 ../
-rw 633317 Mar 6 18:02 Ksecu.img
-rw 1956920 Mar 6 18:01 K2os.img
-rw 18331639 Mar 6 18:02 Kbase.img
-rw 5719825 Mar 6 18:01 Keni.img
-rw 2864354 Mar 6 18:01 Kadvrout.img
-rw 3437 Mar 6 18:01 Kencrypt.img
-rw 792 Mar 6 18:02 software.lsm
-rw 28086 May 16 16:07 boot.cfg

Listing Directory /flash/switch:

drw 1024 Mar 6 18:13 ./
drw 1024 May 21 06:54 ../
-rw 105613 Jun 19 2007 certs.pem
-rw 20 May 21 06:53 snmp.engine
-rw 0 Jun 19 2007 pre_banner.txt
drw 1024 Mar 6 18:13 avlan/
-rw 1732 Mar 6 18:13 label.txt
-rw 973 Dec 18 2007 wv-cert.pem
-rw 963 Dec 18 2007 wv-key.pem

Listing Directory /flash/switch/avlan:

drw 1024 Mar 6 18:13 ./
drw 1024 Mar 6 18:13 ../
-rw 445 Mar 6 18:13 topA.html
-rw 358 Mar 6 18:13 topB.html
-rw 388 Mar 6 18:13 bottomA.html
-rw 384 Mar 6 18:13 bottomB.html
-rw 2444 Mar 6 18:13 myLogo.gif
-rw 8698 Mar 6 18:13 javlan.cab
-rw 13521 Mar 6 18:13 javlan.jar
-rw 1121 Mar 6 18:13 avlanMacInstall.html
-rw 51200 Mar 6 18:13 avlaninstall.tar
-rw 12090 Mar 6 18:13 javlan.jar.sig
-rw 5327 Mar 6 18:13 javlaninstall.sit
-rw 11569 Mar 6 18:13 javlanmacosx.jar
-rw 1722 Mar 6 18:13 reconfigApple.html

Listing Directory /flash/network:

drw 1024 Aug 13 2011 ./
drw 1024 May 21 06:54 ../
-rw 91392 Apr 27 2010 userTable5
-rw 12 Mar 6 16:24 lockoutSetting
-rw 33 Dec 27 2008 policy.cfg
-rw 404 Dec 27 2008 ssh_host_dsa_key
-rw 358 Dec 27 2008 ssh_host_dsa_key.pub

Listing Directory /flash/working:

drw 1024 May 16 15:39 ./
drw 1024 May 21 06:54 ../
-rw 28086 May 16 15:39 boot.cfg
-rw 633317 Mar 6 18:02 Ksecu.img
-rw 1956920 Mar 6 18:01 K2os.img
-rw 18331639 Mar 6 18:02 Kbase.img
-rw 5719825 Mar 6 18:01 Keni.img
-rw 2864354 Mar 6 18:01 Kadvrout.img
-rw 3437 Mar 6 18:01 Kencrypt.img
-rw 792 Mar 6 18:02 software.lsm

4258816 bytes free

Can you tell me if i have to remove files again?

Best Regards,

Bma

[one6f]

The switch booted in certified and when I type this command , there are this message error again!

After image upgrade you should boot in working:

Code: Select all

reload working no rollback-timeout

If you now in certified, try to delete files in working and sinchronyze again to the old release

Code: Select all

show microcode working
show microcode certified
show microcode loaded
rm working/*.img
copy certified working
write memory flash-synchro

Be sure you have enough memory on all NIs:

Code: Select all

rls NI /flash

Check swlog:

Code: Select all

show log swlog timestamp mm/dd/yyyy hh:mm 

Then try to upgrade images to GA release: OS6850&OS6850E_644343R01GA.zip
Check swlog for alerts.
After that:

Code: Select all

reload working no rollback-timeout

Check

Code: Select all

freespace /flash
fsck /flash

Synchronize:

Code: Select all

write memory flash-synchro

Then upgrade uboot/miniboot from GA build.
Reload entire stack from certified.
Check uboot/miniboot version

Code: Select all

show ni
.........
  UBOOT Version :                6.4.4.213.R01,
  UBOOT-miniboot Version :       6.4.4.213.R01,
.........

And only then upgrade to the last maintenance build: OS6850&OS6850E_644502R01.zip

[bma]

One6f,

Thank you very much for your promptness and helpful response.
My problem is solved.

I removed all files .bk from /flash/switch/avlan and the files crash.pmd from /flash from each switch of stack.
Then, I typed the command "write memory", "copy certified working" and copy flash-synchro.
---------------------------------------------------------------
213# copy flash-synchro
Setting CERTIFY Timeout for 800 seconds

CERTIFY process completed successfully
Flash Synchronization process started
+++ == CSM == Stack 1 Certify process Completed
+++ == CSM == Stack 2 Certify process Completed
+++ == CSM == Stack 3 Certify process Completed
+++ == CSM == Stack 4 Certify process Completed
Flash Synchronization process completed successfully

--------------------------------------------------------------------
213# SHOW RUNNING-DIRECTORY

CONFIGURATION STATUS
Running CMM : PRIMARY,
CMM Mode : DUAL CMMs,
Current CMM Slot : 1,
Running configuration : WORKING,
Certify/Restore Status : CERTIFIED
SYNCHRONIZATION STATUS
Flash Between CMMs : SYNCHRONIZED,
Running Configuration : SYNCHRONIZED,
Stacks Reload on Takeover: PRIMARY ONLY
--------------------------------------------------------------
I was surprised to see the running configuration set up automically on working.

I haven't needed to type the command "reload working no rollback-timeout.

Is it usual?


Best regards,

Bma

[one6f]

Is it usual?

Yes, while certified=working working directory takes over the certified.