Experimental Physics and
Industrial Control System

"Jeff Hill" <[email protected]> · Mon, 9 Feb 2009 11:00:41 -0700

Eric,

There is a bug (Mantis 331) related to manual configuration of the beacon
address list on little endian arch (the bug is only in R3.14.10), but I
assume that you are not manually configuring the beacon address list.

> Not all the time, but once in a while, widgets stay white for 3 or 4
> minutes after the IOC has completed iocInit().

If the client was already connected to the IOC and it is rebooted then
connect times should be instantaneous.

Otherwise, if the IOC was turned off for a long time and the client sees
that there is a new beacon, you should not see connect times any longer than
N times five seconds where N is the number of disconnected channels divided
by the number of frames per search attempt times the number of channel
search messages per frame. The number of frames per search attempt is
dynamically adjusted based on past success rates but is a minimum of one.
The number of channel search messages per frame can be approximated by 1024
divided by the average channel name size plus 16 (the protocol header size).

Jeff

> -----Original Message-----
> From: Eric Norum [mailto:[email protected]]
> Sent: Wednesday, February 04, 2009 7:50 AM
> To: Mark Rivers
> Cc: Jeff Hill; Andrew Johnson; Core-Talk
> Subject: Re: Very slow reconnection to medm after IOC reboot
> 
> Just for the record, I see this problem from solaris MEDM sessions to
> R3.14.10 Coldfire/RTEMS IOCs too.
> Not all the time, but once in a while, widgets stay white for 3 or 4
> minutes after the IOC has completed iocInit().
> 
> --
> Eric Norum <[email protected]>
> Advanced Photon Source
> Argonne National Laboratory
> (630) 252-4793

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System