Experimental Physics and
Industrial Control System

Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov> · Wed, 14 Oct 2020 08:34:27 -0700

On 10/14/20 12:33 AM, Abdalla Ahmad via Tech-talk wrote:
> Unfortunately that won't work. Most of these devices have a messed-up hardware clock, the BLM gives a timestamp in 2010, the Libera electron has a timestamp in the 1970 (Still thinking it's Unix Epoch :) ).

It may be worth noting that "/dev/rtc" is not the same thing as the system clock.
(cf. "RTC vs system clock").

https://linux.die.net/man/4/rtc

The RTC is usually only read once at boot time.  NTPD should run just fine without
an RTC.  It might take a bit longer to sync up.  Running "ntpq -p" should allow
you to watch this process.

I was once able to make NTPD work on an embedded linux system without an RTC.
One of the problems I ran into was that the configured tick frequency
(PPC decrementer in this case) initially didn't match the actual frequency.
So it appeared to run too fast.  This was something which I could change
in a device tree file.

> I will try an IOC on a server which contains calc records as a buffer to these PVs, I imagine this will eliminate the timestamp issue.
> 
> Best Regards,
> Abdalla.
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> *From:* Brown, Garth <gwbrown at slac.stanford.edu>
> *Sent:* Tuesday, October 13, 2020 9:20 PM
> *To:* Abdalla Ahmad <Abdalla.Ahmad at sesame.org.jo>; Mark Rivers <rivers at cars.uchicago.edu>; Shankar, Murali <mshankar at slac.stanford.edu>
> *Cc:* tech-talk at aps.anl.gov <tech-talk at aps.anl.gov>
> *Subject:* Re: EPICS timestamp in the archiver appliance
>  
> If you can tolerate a small inaccuracy in the time stamps, and you have PVs on a different device with proper time stamps updating about the same rate, you could work around this by using the .TSEL field. Make the .TSEL field of PVs on the device with inaccurate time a link to the .TIME field of a PV on the device with accurate time, and you'll get time stamps accurate to within the scan rate of the PV on the device with accurate time.
> 
> Garth
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> *From:* Tech-talk <tech-talk-bounces at aps.anl.gov> on behalf of Shankar, Murali via Tech-talk <tech-talk at aps.anl.gov>
> *Sent:* Tuesday, October 13, 2020 7:15 AM
> *To:* Abdalla Ahmad <Abdalla.Ahmad at sesame.org.jo>; Mark Rivers <rivers at cars.uchicago.edu>
> *Cc:* tech-talk at aps.anl.gov <tech-talk at aps.anl.gov>
> *Subject:* Re: EPICS timestamp in the archiver appliance
>  
>>> Is there a way to configure the archiver appliance to take the local timestamp instead of the CA one?
> Sorry; not at this moment. Without a proper timestamp, the archiver appliance will drop samples.
> 
> Regards,
> Murali
> 
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> *From:* Abdalla Ahmad <Abdalla.Ahmad at sesame.org.jo>
> *Sent:* Tuesday, October 13, 2020 5:17 AM
> *To:* Mark Rivers <rivers at cars.uchicago.edu>; Shankar, Murali <mshankar at slac.stanford.edu>
> *Cc:* tech-talk at aps.anl.gov <tech-talk at aps.anl.gov>
> *Subject:* RE: EPICS timestamp in the archiver appliance
>  
> 
> Hi Mark
> 
>  
> 
> No they are something else, specifically they are Libera devices from I-Tech. For example on the Libera electron I managed to get the NTP working by setting the hardware clock using “hwclock”, writing “/etc/ntp.conf” pointing to our NTP server then running the NTP daemon but the Libera BLM for example can’t set the hardware clock because it can’t find “/dev/rtc”.
> 
>  
> 
> We are in talks with I-Tech to solve the issue but I want to know if such configuration can be done on the archiver side. For example, “camonitor –t c” gives you the correct timestamp (local one), so I imagine EPICS somehow can be configured in such manner.
> 
>  
> 
> Thanks!
> 
> Abdalla.
> 
>  
> 
> *From:* Mark Rivers <rivers at cars.uchicago.edu>
> *Sent:* Tuesday, October 13, 2020 3:10 PM
> *To:* Abdalla Ahmad <Abdalla.Ahmad at sesame.org.jo>; 'mshankar at slac.stanford.edu' <mshankar at slac.stanford.edu>
> *Cc:* tech-talk at aps.anl.gov
> *Subject:* RE: EPICS timestamp in the archiver appliance
> 
>  
> 
> Are these hard-IOCs vxWorks, RTEMS, or something else?  On vxWorks you should be able to make NTP work.
> 
>  
> 
> Mark
> 
>  
> 
>  
> 
> *From:* Tech-talk <tech-talk-bounces at aps.anl.gov <mailto:tech-talk-bounces at aps.anl.gov>> *On Behalf Of *Abdalla Ahmad via Tech-talk
> *Sent:* Tuesday, October 13, 2020 7:07 AM
> *To:* tech-talk at aps.anl.gov <mailto:tech-talk at aps.anl.gov>; mshankar at slac.stanford.edu <mailto:mshankar at slac.stanford.edu>
> *Subject:* EPICS timestamp in the archiver appliance
> 
>  
> 
> Hi
> 
>  
> 
> We have an issue when we want to archive PVs from a hard-IOC based device. The issue is that these devices don’t have the time set correctly and most likely somehow NTP won’t work. This causes an issue archiving these PVs because, if I understand correctly, the archiver takes the timestamp from the channel access not the local one. Is there a way to configure the archiver appliance to take the local timestamp instead of the CA one?
> 
>  
> 
> Best Regards,
> 
> Abdalla.
> 

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System