failed to deploy an image ice V2.01

[klemerS]

Hello.

Customer have ice V2.01 - he capture rhel 5U1
from bl460c & try to deploy this image.

He got his first error on the partition page:

Partitioning error:
===========================
Step 4: Specify Partition Information



Partition Information

com.google.gwt.user.client.rpc.InvocationException: The call failed on the server; see server log for details

----------------------------------------------
(btw - the image i captured from the time i build the ice system - works fine )


more logs:

Log:
===========================
ID: 105540
Task Name: Deploy Linux Image
Tool: Deploy Linux Image
Status: Failed

Start time: 4/26/09 3:51 PM
End time: 4/26/09 4:19 PM

Target name: encux5-b9-mg
Target status: Failed

Start time: 4/26/09 3:51 PM
End time: 4/26/09 4:19 PM

--------------------------------------------------------------------------------

Operation 1: Power Down Server

Status: Complete


Start time: 4/26/09 3:51 PM

End time: 4/26/09 3:51 PM

Log:


Checking to see if power is on.
Attempting to power down via management processor
Completed succesfully


Operation 2: Configure System to Boot from Network

Status: Complete


Start time: 4/26/09 3:51 PM

End time: 4/26/09 3:51 PM

Log:


Configuring system to boot from network.
Booting to 64bit ramdisk.
Booting to 64bit ramdisk.
Booting to 64bit ramdisk.
Booting to 64bit ramdisk.
Completed successfully.


Operation 3: Set System to One Time Network Boot

Status: Complete


Start time: 4/26/09 3:51 PM

End time: 4/26/09 3:51 PM

Log:


Setting one time PXE.
Completed successfully.


Operation 4: Power On Server

Status: Complete


Start time: 4/26/09 3:51 PM

End time: 4/26/09 3:52 PM

Log:


Checking to see if power is on.
Completed succesfully


Operation 5: Wait for RAM Disk Response

Status: Complete


Start time: 4/26/09 3:52 PM

End time: 4/26/09 3:59 PM

Log:


Waiting for RAM disk to respond.
RAM disk booted sucessfully at 10.57.18.132


Operation 6: Deploy Linux Image

Status: Failed


Start time: 4/26/09 3:59 PM

End time: 4/26/09 4:19 PM

Log:


Configuring partition scheme to send to system.
Configuring RAM disk for deployment.


Deployment Commands


Waiting for RAM disk to complete configuration.
Ramdisk did not respond in allotted time


Operation 7: Configure System to Boot from Local Disk

Status: Not Executed


Start time: 4/26/09 4:19 PM

End time: 4/26/09 4:19 PM

Log:


This operation has a status of Not Executed because a previous operation in the chain has a status
of Failed.


Operation 8: Reset Server

Status: Not Executed


Start time: 4/26/09 4:19 PM

End time: 4/26/09 4:19 PM

Log:


This operation has a status of Not Executed because a previous operation in the chain has a status
of Failed.





Any idea what could be the problem ???


regards


Shalom Klemer.

[Mitchell Kulberg · Accepted solution]

Hello,

There is a known problem in V2.01 where an image that takes longer than 30 minutes to deploy fails due to a timeout.

This is discussed in an earlier thread on this forum at:

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1268099

If you contact support, they should have a patch for you to correct the problem. If you can't get it, let me know.

Mitch

[klemerS]

Thank for replay.

I will open a case And ask for the PATCH.


Regards.

Shalom.


BTW. - why i got on the partition page the error ,& not the possebility to view or change it ??

[Mark Schroeder]

There is another possibility. Were the capture and the deploy both done using the same version of ICE-Linux? I.e., V2.01?

[klemerS]

Yes. It is the same version.

I build the system a year ago with V2.01

Regards

Shalom.

[Rigoberto Corujo]

Hi Shalom,

As far as the error on the partition page, I've seen that happen when you press "Next" too fast before the page has fully loaded.

Rigoberto

[Mark Schroeder]

If it is repeatable, it would help to know what your partition scheme looks (looked) like.
If the system is available, then 'sfdisk -l' output would help. If not, then /opt/repository/.icle/.xml has the partition information, starting with <Storage> and ending with </Storage>

[klemerS]

I opened a call. I will be at customer site on thursday , I hope i will get the patch till then. I will do all checks you suggested guys.

Thank you.

Shalom

[klemerS]

Hello again.

Very strange !!!

/var/vgcrash which is on san disk , size 40gb , when looking into xml file it goes into same partition as vg00. ( c0d0p2 )& sdr.

I am sending the xml file + the relevant entry.

<vg name="vg00" size="61407232">
<pvdevice partindex="2" name="disk0">/dev/cciss/c0d0p2</pvdevice>
<lv name="lvOnGuard" size="524288" device="/dev/vg00/lvOnGuard" type="linux">
<fs>ext3</fs>
<uuid>5f03eba0-2abe-401b-bec7-43dc3cd1a591</uuid>
</lv>
<lv name="lvprecise" size="3145728" device="/dev/vg00/lvprecise" type="linux">
<fs>ext3</fs>
<uuid>049c1547-5a7a-4c06-a0ae-4e7ff62aba30</uuid>
</lv>
<lv name="lvol1" size="4194304" device="/dev/vg00/lvol1" type="linux">
<fs>ext3</fs>
<uuid>086e78b5-a12e-480e-a302-d7fad21e0a6b</uuid>
</lv>
<lv name="lvol2" size="4194304" device="/dev/vg00/lvol2" type="linux">
<fs>ext3</fs>
<uuid>32c14cf0-d1e2-4539-a5d7-fde9e78a1cfa</uuid>
</lv>
<lv name="lvol0" size="5144576" device="/dev/vg00/lvol0" type="linux">
<fs>ext3</fs>
<label>/opt</label>
<uuid>d5ccc931-6529-43d1-91ca-922241807a58</uuid>
</lv>
<lv name="lvol3" size="5144576" device="/dev/vg00/lvol3" type="linux">
<fs>ext3</fs>
<uuid>ece66437-6a5a-48db-9aca-3f949c440c5a</uuid>
</lv>
<lv name="lvOracleTools" size="15728640" device="/dev/vg00/lvOracleTools" type="linux">
<fs>ext3</fs>
<uuid>047774b8-3dd3-4c34-9e68-dd5dc1640fb7</uuid>
</lv>
<lv name="lvswap" size="*" device="/dev/vg00/lvswap" type="swap">
<fs>swap</fs>
</lv>
</vg>
<vg name="vgcrash" size="94347264">
<pvdevice partindex="2" name="disk0">/dev/cciss/c0d0p2</pvdevice>
<pvdevice>/dev/sdr</pvdevice>
<lv name="lvcrash" size="*" device="/dev/vgcrash/lvcrash" type="linux">
<fs>ext3</fs>
<uuid>1d6d062b-5e47-4a4f-81f0-7305fedfcafd</uuid>
</lv>
</vg>


Best Regards.

Shalom.

[Mark Schroeder]

We're pretty surprised that you show two "pvdevice" entries for vgcrash. We're investigating whether that makes sense and if not, how it could occur.

[klemerS]

Thanks.

I can open a new case & send you sosreport if you wish. we tried the capture 3 times with same results. we even export vgcrash but got same results.

we tried to delete all none relevant disks from the xml file including vgcrash & on the partition page we got the partition table.
we saw as mountpoint only / + /opt .all other lvols were without mountpoint.

the Deploy start o.k. but failed later

regards

Shalom.

[Mark Schroeder]

It sounds as if removing the (odd) partitions with two "pvdevice"s at least got the partition wizard GUI to load ok. Correct? Then some other volumes were missing mount points in the partition wizard and the deploy failed. Also correct? (Just checking my understanding.) If so, you've probably hit two independent defects.

Can you try this?

* Remove the non-essential (odd) partition with the two "pvdevice"s. This should allow the partition wizard GUI to load properly.

* For any partition in the wizard that does not show a mount point, edit that partition and set its mount point. Each partition must have a mount point set, but the partition wizard incorrectly assumes that the partition "label" contains the mount point and that is not always the case. This issue has recently been release noted. Setting the mount point by hand by editing the partition information is the workaround.

* Once all partitions have mount points set in the partition wizard, "continue" in the partition wizard and perform the deploy.

I would like to know if the above steps result in a successful deploy.

[klemerS]

O.K. thanks !!!!!!

I just sent what you asked to customer.

hoping for quick answer.

Regards

Shalom.

[Mark Schroeder]

This has remained open for some time. Perhaps the workaround solved your problem. I am still concerned that you had
a bad capture (two pvdevice entries).

What can we do to reproduce the bad capture? Do you understand what about
your configuration is causing that to happen?

[klemerS]

Hello.

I am sorry for the delay.

i wasnt working for 3 weeks.

Customer told me he cant tests the systems since its goes to production.

he will do some tests on other systems in the next 2 weeks.

I will soon close the case.

Thank for your help !!!!

[klemerS]

Hello Again.

Our Error happened again with other system.

this system connect to EMC too , but the OS is on local disks.

We did what Mark suggested & its worked. The deploy works fine on one system & failed on another , But from the deploy error on second bl460c its seems like a known problem , which V2.10 solves.

"aer : probe of 000xxx:pcie01 failed with error 2"


I opened a case regarding the double PV. I will send the image xml file & sosreport from source bl460c.


Best Regards.

Shalom.

[klemerS]

hi.

Sorry for the Delay.

i got an answer , to upgrade the ICE version to solve the problem.
thanks to all.

Shalom.