RE: SNMP device support

[<nick.rees@diamond.ac.uk>]

I believe all snmp variants derive (very loosely) from the same source.

Sheng made did a major re-write of the DESY code (for LCLS, I believe) at some point between 2006 and 2009. It looks like NSCL/FRIB did too, and when I looked at this in 2009 Euan Troup from CSIRO and Janet Anderson from APS had also had a go. I wasn't aware of the NSCL/FRIB developments at the time.

So, when I had problems with the DESY driver back in 2009, Sheng sent me his modifications and of all the work I saw at the time it was the best developed. I folded his changes back into the DESY distribution and forwarded it back to DESY in 2009. The last message I had was from Matthias Clausen on July 2009 saying that they would fold it in and release it "next week". It appears that didn't happen.

We have since patched the code for the memory leak (see attached patch file), but since support seems to have died at DESY, we should decide on a way forwards. I haven't done much snmp development here (just APC power strips, just like Marty). It looks like John Priller's rewrite is even bigger than Sheng Peng's - it is now largely C++ and unrecognisable from the original, but it has more functionality than Sheng's and, in particular, supports SNMP writes.

I haven't tested the FRIB code yet, but it compiles OK and it does seem that the NSCL/FRIB dependency on SNMP is greater than ours, so I would hope that indicates a better level of support. Hence, I am looking at changing to the NSCL/FRIB driver and would cautiously recommend you to do so too. 

Cheers,

Nick Rees
Principal Software Engineer           Phone: +44 (0)1235-778430
Diamond Light Source                  Fax:   +44 (0)1235-446713

> -----Original Message-----
> From: tech-talk-bounces@aps.anl.gov [mailto:tech-talk-
> bounces@aps.anl.gov] On Behalf Of Marty Smith
> Sent: 19 March 2013 10:58
> To: Florian Feldbauer
> Cc: EPICS Tech Talk
> Subject: Re: SNMP device support
> 
> Hi Florian,
> 
> I have been testing an application using the driver from Sheng Peng
> since it does provide write capabilities via SNMP. However, I have
> noticed that during my testing the on a Linux-x86 64-bit machine
> running under EPICS
> R3.14.12.2 that the soft IOC does use more and more memory over time
> indicating a possible memory leak. However, I have not confirmed the
> source of the problem at this time. I am trying to find some time to
> test the one from John Priller but as of yet have not been able to.
> 
> My application is providing data from rack mounted APC power strips. In
> the next few weeks I intend on finding some time to set up a test using
> John's driver though which Eric Norume has provided updates for.
> https://groups.nscl.msu.edu/controls/files/snmp-nscl-1.0.RC6.tgz
> 
> Marty Smith
> ANL/APS Controls Group
> 
> 
> ----- Original Message -----
> From: "Florian Feldbauer" <florian@ep1.ruhr-uni-bochum.de>
> To: "EPICS Tech Talk" <tech-talk@aps.anl.gov>
> Sent: Tuesday, March 19, 2013 5:19:12 AM
> Subject: SNMP device support
> 
> Dear all,
> 
> for our low voltage power supply, we are thinking about using the
> Wiener MPod which is controlled via SNMP.
>  From Wiener I know that there are some implementations for EPICS:
> - The original devSNMP from DESY
> http://www-mks2.desy.de/content/e4/e40/e41/e12212/index_ger.html
> - An extension of the first one from John Priller
> http://groups.nscl.msu.edu/controls/
> - and a third one from Sheng Peng from Diamond light source
> http://controls.diamond.ac.uk/downloads/misc/snmp.tgz
> 
> Is there any great difference in those device supports? Especially
> between the second and third one?
> Are there any recommendation which one to use?
> 
> Best regards,
> Florian
> 
> --
> ----------------------------------------
> | Dr. Florian Feldbauer                |
> |                                      |
> | Helmholtz-Institut Mainz /           |
> | Johannes Gutenberg-Universität Mainz |
> | Johann-Joachim-Becher-Weg 45         |
> | D-55128 Mainz                        |
> |                                      |
> | Office: 01/217                       |
> | Phone:  (+49)6131 / 39-25829         |
> ----------------------------------------
>

Attachment: snmp_leak.patch
Description: snmp_leak.patch