HIMANSHU schrieb:
> IET target is connected by MS initiator.Data transfer is going on on
> the connected disk on initiator.

I don't see how your question fits here - you use IET as a target, and a 
MS initiator.
This is open-iscsi (initiator) list, not a general 
iSCSI-related-questions list.

> TCP connection failed(because of any reason,service failure,problems
> in kernel modules)  at say 't1' second.Initiator kept trying for
> reconnect.
> Initiator actually recognizes that target is down at say 't2' seconds.
> Then for the time between 't1' and 't2',data is written on the cache
> on initiator side giving illusion of data transfer.After some time,it
> gave delayed write error.
> So data is not actually stored on the target.
> What can we do here?Using high capacity network can be the solution in
> case of dedicated framework so that initiator will recognize the
> failure as early as possible.

open-iscsi has configurable timeouts.

For example, by setting "node.session.timeo.replacement_timeout" to a 
higher value than the default 120 seconds allows your initiator to 
"survive" network failures lasting minutes, hours, days etc., depending 
on how big the timeout you set.
Any task which needs to access data on the target, but can't, because of 
a network failure, will be put into an "uninterruptible sleep" until the 
target is accessible again (or the timeout triggers, which means I/O error).

Tomasz Chmielewski

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 
For more options, visit this group at http://groups.google.com/group/open-iscsi

Reply via email to