Re: deny ca_search response for a record?

[Paul Sichta <psichta@pppl.gov>]

Andrew,

I have been able to change the ASG field of a record at runtime, so by doing this and adding a "no r/w"  rule to my ASG file I should be able to dynamically set no-read condition to a record.

As Andrew points out,  I'll have to evaluate caget's response to the no-read condition as opposed to not-found.

Thanks,

-ps

On Fri, Apr 29, 2016 at 1:42 PM, Andrew Johnson <anj@aps.anl.gov> wrote:

Hi Paul,

Using Channel Access security to deny read access will give you one kind
of response which is worth testing for, but it won't be quite the same
as what happens if the network connection to the IOC dies and then gets
restored. Unreadable channels can still be found, you just won't get any
data from or about them, so making a channel unreadable isn't the same
as the server completely disconnecting.

You can send a Unix IOC process the SIGSTOP and SIGCONT signals to
suspend/resume it (or press Crtl-Z on the console to stop, run fg to
resume). Somewhat more complicated might be to enable and disable a
firewall rule to block network packets between the client and server.

- Andrew

On 04/29/2016 12:21 PM, Paul Sichta wrote:
> Kay,
> That's a good solution, and I can load "test support"  .ascf  files
> without rebooting the IOC. The record I want to affect uses an ASG that
> other records also have, so I'll have to see if I can set precedence in
> the .ascf  file  to single out just that one record for "no read".
>
> Thanks,
> -ps
>
>
> On Fri, Apr 29, 2016 at 12:48 PM, Kasemir, Kay <kasemirk@ornl.gov
> <mailto:kasemirk@ornl.gov>> wrote:
>
>     Hi Paul:
>
>
>     I believe the channel access rules security not only cover write
>     access but also read access.
>
>
>     So you could have a channel access security rules which, based on
>     the value of some PV, allows or prohibits read access to your test
>     records.
>
>
>     -Kay
>
>
>     ------------------------------------------------------------------------
>     *From:* tech-talk-bounces@aps.anl.gov
>     <mailto:tech-talk-bounces@aps.anl.gov>
>     <tech-talk-bounces@aps.anl.gov
>     <mailto:tech-talk-bounces@aps.anl.gov>> on behalf of Paul Sichta
>     <psichta@pppl.gov <mailto:psichta@pppl.gov>>
>     *Sent:* Friday, April 29, 2016 12:11 PM
>     *To:* tech-talk@aps.anl.gov <mailto:tech-talk@aps.anl.gov>
>     *Subject:* deny ca_search response for a record?

>
>     For testing a script with caget's in it I'd like to test it's
>     response to a "timeout" situation, in a transient fashion to test
>     error handling/recovery.  Is there a way from the IOC console to
>     cause a record to act "hidden", and then make it normal again?
>     dbpf into the .NAME field?
>
>     I realize I can add logic in the script itself to simulate this
>     condition, but was just wondering about the alternative.  I also
>     have a fuzzy feeling I can use a gateway to do something like this
>
>     -ps
>
>
>
>

--
There are only two hard problems in distributed systems:
  2. Exactly-once delivery
  1. Guaranteed order of messages
  2. Exactly-once delivery
 -- Mathias Verraes