EPICS Controls Argonne National Laboratory

Experimental Physics and
Industrial Control System

2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  <20162017  2018  2019  2020  2021  2022  2023  2024  Index 2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  <20162017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: Re: EPICS 3.15.5 planning
From: Andrew Johnson <[email protected]>
To: <[email protected]>
Date: Mon, 3 Oct 2016 11:23:32 -0500
Hi Ralph,

On 10/03/2016 09:29 AM, Ralph Lange wrote:
> I would like to push out a Base 3.15.5 release in time for our upcoming
> CODAC Core System 5.4 update.
> 
> My must-have list currently has two items:
> 
>  1. Fix for scalability issue when loading many records
>  2. Add missing other-time-to-epics-time conversion without timezone
>     mechanism
> 
> Both fixes should probably be applied to 3.14 and merged up. Right?

Is your scalability fix plan to apply Michael's bigdb branch?
    https://code.launchpad.net/~epics-core/epics-base/bigdb
I just added another review comment to that branch, it needs
dbCreateAlias() to be updated too before it can be merged.

The speed-up is achieved by sorting the records by name at the end of
each call to dbLoadRecords() instead of the current insertion sort. We
could speed things up even more by not keeping the records in sorted
order at all. Users might notice the difference in the output order from
the reporting tools that scan the whole database, but they don't get a
full sort even now because the sorting only occurs within each record
type. Making this additional change on the 3.14 branch would complicate
matters though, because 3.14 still builds the dbStaticHost library so we
would have to fix dbRenameRecord() as well (but we wouldn't need the
ellLib changes).

You might want to measure how long it takes your 175K database to do
that final sort to decide whether the final-sort is worth keeping (I
hope you're already calling dbPvdTableSize(65536) in that IOC's start-up
script). We could also make that final-sort conditional on a
configuration variable.

Please publish a branch for your fix #2 ASAP.

> Are there any other immanent issues/features that I should target to
> include?

I started merging the 3.14 fixes into 3.15 last week, but I haven't
actually committed that merge to Launchpad yet. I'll try to finish that
off today — I think it's just about ready to go.

> I would like to get this done asap, aiming for end of this week.
> (Tagging is enough for pulling my clone. Website, announcement et al.
> are a lot more relaxed.)
> Comments, screams, advice?

Not impossible; I will be on vacation this Friday and all next week.

- Andrew

-- 
Arguing for surveillance because you have nothing to hide is no
different than making the claim, "I don't care about freedom of
speech because I have nothing to say." -- Edward Snowdon

References:
EPICS 3.15.5 planning Ralph Lange

Navigate by Date:
Prev: EPICS 3.15.5 planning Ralph Lange
Next: Build failed in Jenkins: epics-base-3.14-win32s #145 APS Jenkins
Index: 2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  <20162017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: EPICS 3.15.5 planning Ralph Lange
Next: Build failed in Jenkins: epics-base-3.14-win32s #145 APS Jenkins
Index: 2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  2015  <20162017  2018  2019  2020  2021  2022  2023  2024 
ANJ, 04 Oct 2016 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·