Experimental Physics and
Industrial Control System

Ben Franksen via Core-talk <core-talk at aps.anl.gov> · Tue, 24 Mar 2020 18:02:59 +0100

Hi Ralph

I have thought a bit about this.

What if I want my database to process some counter or debug record via
CPP input link? I may not care at all for the value, only the fact that
an event happens. IMO this is valid use case and it will be hard to make
it work reliable if events are dropped in this way.

BTW:

My initial refactor now passes all tests and is cleaned up and
committed. I have pushed it to lp:~bfrk/+git/epics_base on branch
dbChannelForDbPut (stupid name, sorry).

I tried to issue a merge request for branch 7.0 of lp:epics-base but I
get a failure message saying:

lp:~bfrk/+git/epics_base is not mergeable into this repository.

I have no idea how to fix that. I guess i have been doing something
wrong on the git side. My branch is based on Dirk's proposed branch
dbChannelForDBLinks, perhaps this is what causes the problem?

Cheers
Ben

Am 23.03.20 um 20:04 schrieb Ben Franksen via Core-talk:
> Am 23.03.20 um 12:10 schrieb Ralph Lange:
>> On Sun, 22 Mar 2020 at 11:13, Ben Franksen <
>> benjamin.franksen at helmholtz-berlin.de> wrote:
>>
>>> [...]
>>> Here is the the first thing I am seriously stumbling over:
>>>
>>> static void db_queue_event_log (evSubscrip *pevent, db_field_log *pLog)
>>> {
>>>     ...
>>>     /*
>>>      * if we have an event on the queue and both the last
>>>      * event on the queue and the current event are emtpy
>>>      * (i.e. of type dbfl_type_rec), simply ignore duplicate
>>>      * events (saving empty events serves no purpose)
>>>      */
>>>     if (pevent->npend > 0u &&
>>>         (*pevent->pLastLog)->type == dbfl_type_rec &&
>>>         pLog->type == dbfl_type_rec) {
>>>         db_delete_field_log(pLog);
>>>         UNLOCKEVQUE (ev_que);
>>>         return;
>>>     }
>>>     ...
>>> }
>>>
>>> I don't quite get the logic here. What is (the definition of) an empty
>>> event? And why does dbfl_type_rec indicate that the event is empty?
>>>
>>
>> Probably a bad term.
>
> No, with your explanation below is makes a lot of sense.
>
>> A dbfl_type_rec update does not contain any data or metadata by itself -
>> that's why I called it "empty". It's just a placeholder, signaling to the
>> post-queue processing that it should use the current state and data of the
>> record, including all metadata and timestamp.
>> Putting two consecutive dbfl_type_rec events on the queue would lead to an
>> identical update (as all data is taken from the record) being sent twice.
>> Not really useful, that's why I decided it to be safe to drop any
>> consecutive dbfl_type_rec updates.
>
> I see. In my current version dbfl_type_rec no longer exists, which is
> why I have, for the moment, disabled this optimization. I think I could
> recover it, at least partly, but I am not sure it is worth the trouble.
> This will probably come up again during review.
>
> Cheers
> Ben
>
> ________________________________
>
> Helmholtz-Zentrum Berlin für Materialien und Energie GmbH
>
> Mitglied der Hermann von Helmholtz-Gemeinschaft Deutscher Forschungszentren e.V.
>
> Aufsichtsrat: Vorsitzender Dr. Volkmar Dietz, stv. Vorsitzende Dr. Jutta Koch-Unterseher
> Geschäftsführung: Prof. Dr. Bernd Rech (Sprecher), Prof. Dr. Jan Lüning, Thomas Frederking
>
> Sitz Berlin, AG Charlottenburg, 89 HRB 5583
>
> Postadresse:
> Hahn-Meitner-Platz 1
> D-14109 Berlin
>

________________________________

Helmholtz-Zentrum Berlin für Materialien und Energie GmbH

Mitglied der Hermann von Helmholtz-Gemeinschaft Deutscher Forschungszentren e.V.

Aufsichtsrat: Vorsitzender Dr. Volkmar Dietz, stv. Vorsitzende Dr. Jutta Koch-Unterseher
Geschäftsführung: Prof. Dr. Bernd Rech (Sprecher), Prof. Dr. Jan Lüning, Thomas Frederking

Sitz Berlin, AG Charlottenburg, 89 HRB 5583

Postadresse:
Hahn-Meitner-Platz 1
D-14109 Berlin

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System