Konrad Rzeszutek schrieb:
>> "Synchronizing SCSI cache for disk" happens because:
>>
>> - iSCSI sessions were not properly disconnected, and
> 
> Correct.
> 
>> - they can't be properly disconnected any more, because the network is 
>> already disabled.
> 
> Kind of. There is a kernel timer that gets activated during the logout 
> sequence
> that waits for up to 120 seconds (or what you have set in
> node.session.timeo.replacement_timeout) and if the logout sequence hasn't 
> completed releases the kernel resources.
> 
>> Most distributions shut down all network interfaces when a "halt" 
>> command is started (i.e., they add "-i" option to the halt command):
>>
>>      -i: shut down all network interfaces.
>>
>> Without this flag, everything should shut down properly, even when it's 
> 
> Right. And this situation will hang the kernel during reboot b/c the
> SCSI error handlers wait for a logout state condition that never happens.
> 
>> not possible to logout all sessions earlier (i.e., a diskless machine 
>> started off iSCSI).
> 
> And the patch I attached in the previous e-mail describes a solution
> to this.

BTW, similar hack (not disabling the network) is also needed when we 
reboot the system using kexec (without that patch, that is):

  -x, --no-ifdown      Don't bring down network interfaces.


-- 
Tomasz Chmielewski
http://wpkg.org

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To post to this group, send email to open-iscsi@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/open-iscsi
-~----------~----~----~----~------~----~------~--~---

Reply via email to