| 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 <2022> 2023 2024 2025 | Index | 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 <2022> 2023 2024 2025 |
| <== Date ==> | <== Thread ==> |
|---|
| Subject: | Re: Problem with huge waveforms in EPICS 7 |
| From: | Michael Davidsaver via Core-talk <core-talk at aps.anl.gov> |
| To: | "Zimoch Dirk (PSI)" <dirk.zimoch at psi.ch> |
| Cc: | "core-talk at aps.anl.gov" <core-talk at aps.anl.gov> |
| Date: | Thu, 30 Jun 2022 03:35:00 -0700 |
On 6/29/22 23:59, Zimoch Dirk (PSI) wrote:
On Wed, 2022-06-29 at 10:34 -0700, Michael Davidsaver wrote:On 6/28/22 00:13, Zimoch Dirk (PSI) wrote:Confirmed with my test setup.Applied as https://github.com/epics-base/epics-base/commit/4df48c91f4e7202899bd859312a2b68148bf0bad I also made a note on the PR https://github.com/epics-base/epics-base/pull/99#discussion_r910230977Thanks! This issue has shown one of the limitations of the ci tests as we have them now: The problem only shows when ioc and client run on two different hosts, so that network bandwidth is an issue. Is there any way to cover such cases in the test environments?
A more tractable problem would be testing the overflow behavior of dbEvent. I think this could be accomplished by causing a callback set via db_add_extra_labor_event() to block on an epicsEvent to prevent the queue from emptying. Then post events until the queue overflows. Then trigger the epicsEvent, and allow the queue to empty.