Incorrect HASH

[Tuomas järvinen]

Hi there!

 

Having huge problem with OS deploy

after downloading WIM file, ts manager stops for incorrect hash 0x80091007 error.

 

I have tested lot of things, but still hope, that I don't need to reinstall production environment.

Image is valid, removed, updated, verified, scanned, hash remains same (using hashdir.exe, md5 or sha checksuming tools) in source and DP, but after TS manager copies it by http, hash is no more same.

 

This happens, even I have created brand-new image....

 

 

For reference, http://social.technet.microsoft.com/Forums/en-US/8599b710-65e6-46a8-a462-a505e735cc52/issues-with-osd-0x80091007-incorrect-hash?forum=configmanagerosd

 

unless someone has something .....

 

~T

[Peter van der Woude]

Are you using any antivirus/antimalware client on the distribution point server, or on the server hosting the source directory?

[Tuomas järvinen]

I have uninstalled all security products from DP and site servers ealier.

 

Also re-installed the DP role, but did not helped for this problem.

[anyweb]

have you tried to redistribute/revalidate the package on that dp where it's generating the hash error ? what does the smsdpmon.log tell you about validation of that package ?

[Tuomas järvinen]

Yes I have - as I told in ms forum side.

 

Before friday, smsdpmon.log shows "Failed to find algorighm ID from registry. Use default algorithm" error. However, after re-installing DP role it disappeared. I updating or validating package does not show any additional message. I can use hashdir (or other equivalent) tool to verify that .WIM file is same in package source and DP, but after it is transferred to client in WinPE, (and copied to .another pc since x64 winpe can not execute 32bit executables), checsu is not equal. File size still equals...

 

I can of course use package share options, but then process might stop when downloading drivers...

 

~T

[anyweb]

do you have any QOS software/hardware on your network that could be causing issues ? have you checked if the switches inbetween the dp and the clients are ok ?

[Tuomas järvinen]

Nothing I could mention. Disabled (Microsoft) QOS Packet scheduler on adapters properties if it interfers somehow.

 

Asked our network team to inspect switch in differnet offices Theyt said, some single packet drops ( of 10 billion packets), but happens also there, where image trasnfer is Ok.

[Tuomas järvinen]

Managed to solve the problem with an expert.

It seems that there was a kind of "blockage". Couple packages was flagged with "progress" status on Distribution point configuration status -> DP -> details tab. After package redistribution, the image was transferred and applied fine.

 

I have validated all content on every DP at last weekend and redistributed some packages showing error. However summary tab still shows warning

and there is only one message (more than one month old) stating: """" Failed to retrieve the package list on the distribution point ["display=\xxDP1.fqdn.lan"]MSWNET:["SMS_SITE=zzz"]\\xxDP.fqdn.lan\ or the package list in content library doesn't match the one in WMI. Review smsdpmon.log for more information about this failure."""

 

Had to say that I did not remember this kind of situation before R2 (or CU1/CU2).

 

~T

[anyweb]

as you said the operating system image was failing with hash errors did you do as I advised ?

 

have you tried to redistribute/revalidate the package on that dp where it's generating the hash error ?

 

 

both of those actions should have alerted you to problems with the package as would monitoring the smsdpmon.log

 

same goes for any other package with an issue on the affected dp

 

glad it's working now

[Tuomas järvinen]

I know. Too much to do and still so much time to monitor SCCM environment....

 

But still, thanks.

~T