HP Flex10 replacement firmware update

标签:
it |
分类: 工作 |
One of the flex10 module failed and got replacement from HP. It taking a long time to update its firmware, detail story as below.
Our
environment:
2 HP 7000 chassis in one domain for the virtual connector. (Primary
chassis OA1 and secondary chassis OA2)
Issue:
Port 1 failed to communicate from OA2 interconnect Bay1. Called HP
and have the flex10 module replaced.
The replacement come with firmware 2.2 and all our other bays running firmware 3.30. update is needed for the new replacement.
First Try
(with HP vcsu V1.5.4):
Replaced the new one to OA2 bay1 and try update it, failed for the
following reason:
[Thu Nov 10 14:47:35 2011] (7628)[VcsuCore.cpp:2092] : Primary VCM
found in enclosure USE11695LX, bay 1
[Thu Nov 10 14:47:35 2011] (7628)[VcsuUpdate.cpp:1568] : Deleting
FC firmware images from the Enet modules
[Thu Nov 10 14:47:46 2011] (7628)[VcsuUpdate.cpp:1579] : VCFC image
deletion from bay 1 failed
[Thu Nov 10 14:47:56 2011] (7628)[Sftp.cpp:57] : Deleting remote
file: /images/archives/vcfc2/SWBD82/v6.1.0_uth5
[Thu Nov 10 14:47:56 2011] (7628)[Sftp.cpp:63] : sftp_unlink
failed, errorno = 0
[Thu Nov 10 14:47:56 2011] (7628)[Sftp.cpp:1115] : Delete file
failed
[Thu Nov 10 14:47:56 2011] (7628)[VcsuUpdate.cpp:1604] : VCFC2
image deletion from bay 2 failed
[Thu Nov 10 14:47:56 2011] (7628)[VcsuProgress.cpp:156] : Starting
Activation
[Thu Nov 10 14:47:56 2011] (7628)[VcsuProgress.cpp:796] :
Activation order is ODD EVEN
[Thu Nov 10 14:47:56 2011] (7628)[VcsuProgress.cpp:801] :
Activation for single bay update
[Thu Nov 10 14:47:56 2011] (7628)[VcsuProgress.cpp:819] : Primary
bay number passed for -b option. Exiting activation process since
update was not success
[Thu Nov 10 14:47:56 2011] (7628)[VcsuUpdate.cpp:653] : Firmware
update failed on one or more modules
Second Try
(with HP vcsu V1.6.0)
I thought that maybe is the version of vcsu is too old for this, so
download the new one and try to update it again. Failed! Here is
the log:
[Thu Nov 10 15:38:34 2011] (5836)[VcsuCore.cpp:1955] : Primary VCM
found in enclosure USE11695LX, bay 1
[Thu Nov 10 15:38:44 2011] (5836)[VcsuSftp.cpp:563] : SFTP open
failed for /images/archives/vcmConfig.tgz: errno = 11, Resource
temporarily unavailable
[Thu Nov 10 15:38:44 2011] (5836)[VcsuCore.cpp:4143] : Download of
VCM config file failed
[Thu Nov 10 15:38:44 2011] (5836)[VcsuInit.cpp:891] : Failed to
read generation numbers from VCM
[Thu Nov 10 15:38:44 2011] (5836)[AuthenticateUser.cpp:135] :
authenticateVCM failed
Third and fourth try was soft/hard rest the module and it still cannot update.
CAlled HP
for support and been told that try to put the new module to the OA2
and the bay other than Bay1. it could be the Active Bay and cuase
the update failed. ( what is that about ???) Anyway, take the word
and give it fifth try. and this time even worse.
[Fri Nov 11 09:56:37 2011] (3492)[VcsuCore.cpp:2126] : Module not
present in bay 8
[Fri Nov 11 09:56:38 2011] (3492)[VcsuCore.cpp:1146] : Domain
configuration is present
[Fri Nov 11 09:56:38 2011] (3492)[VcsuCore.cpp:1146] : Domain
configuration is present
[Fri Nov 11 09:56:39 2011] (3492)[VcsuCore.cpp:1146] : Domain
configuration is present
[Fri Nov 11 09:56:40 2011] (3492)[VcsuCore.cpp:1146] : Domain
configuration is present
[Fri Nov 11 09:56:41 2011] (3492)[VcsuCore.cpp:1176] : This is a
subordinate enclosure
[Fri Nov 11 09:56:41 2011] (3492)[VcsuInit.cpp:744] : Primary OA is
not in this enclosure. Getting primary OA IP
[Fri Nov 11 09:56:52 2011] (3492)[VcsuInit.cpp:765] : VCM user
authenticated successfully
[Fri Nov 11 09:57:03 2011] (3492)[VcsuInit.cpp:778] : Unable to get
primary OA IP
Well, I guess I need to make another call to HP, and it come out with this scenario:
Place the new module to the OA1 empty bay and try it. finally, it worked. But it something scary? what if I do not have empty space in the primary chassis? and also, the HP support would suggest to use vcsu 1.5.4 instead of 1.6.0.
Lesson learned from HP, it is not that smart for the chassis blade
system and there is a potential to stop all service in this kind of
environment. I wish they can have more stable and smart way to deal
with this. maybe next 2 version of the firmware or vcsu can solve
it. What a pain!!!