Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Fri, 06 Sep 2002 12:04:34 -0500

Andreas Luedeke wrote:

You can corrupt the record name in the process database.
To "disconnect" the PV "myPV" just generate the script
a=dbNameToAddr("myPV")
**a='@'
and run it on your IOC.
The channel will be invisible for channel access.
If you provide the same channel name now by
another IOC, all applications will use this new channel.

While this might appear to work, I would *strongly* discourage it, as it
will at the least cause confusion and could potentially cause damage to
your facility. The main problem is that it won't make CA clients or
database links that are already connected to the record switch to the new
IOC. For long-running CA clients like the Alarm handler that could be
disasterous (what are the consequences of an alarm from the new record not
being seen for several hours or even days if someone forgets to restart ALH?).
If you change the record's name field like this, you'll almost never be
able to connect to it with the new name unless the hash value of the new
name is the same as that of the old one (ie you can't use the above to
rename records at runtime. It should be possible to at least fix that
problem like this:
  ent=dbAllocEntry(pdbbase)
  dbFindRecord(ent, "myPV")
  dbRenameRecord(ent, "deadPV")
  dbFreeEntry(ent)
However even this approach won't reconnect any live CA clients or links
that point to the record, and it might wake up some other nasty gremlins
that I haven't forseen.
- Andrew
--
"Life is what happens while you're busy making other plans."
-- John Lennon

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System