Here is a few things first:

This will not help you when the host is down or powered off, it will only help you when a service (STORE_SCP,QUERY_SCP, etc.) is stopped. Use the latest version, it runs on both UNIX and Windows.

What it does:

Suppose you have a listener bound to port 104 (NEMA) that is servicing STORE_SCP. You need to update the software on that particular server and will need to take down the listener.

Fine, enter rinetd.

Once your service is stopped, killed, or broken, whip up a quick configuration file…

c:\>cat c:\rinetd.conf

#thishostip 104  anotherhostip 104
192.168.23.4 104 192.168.23.6 104

logcommon
logfile c:\temp\rinetd.log

execute rinetd with the following parameters…

c:\>rinetd.exe -c “c:\rinetd.conf” 

Ok, so what did that buy you? Well, you are now redirecting any DICOM that may come flying at that server while it was down to another listener on another box. This way you don’t have to monkey with the modality to reconfigure it to a secondary or tertiary send destination for maintenance and tech’s can continue their business without even knowing anything is going on…

You can just bounce it. Cool eh? boing boing boing boing…

Tail your log file for evidence its working and tell your friends how zer0-day you are for patching the system without bothering any techs and risking any send failures.

Notable here is that you should add your bounce host as a valid sending host, the redirect does not come from the source, but rather from where it is bounced. This can be problematic if you do any back end disco with your importers that is identified by source.