Back Midas Rome Roody Rootana
  Midas DAQ System, Page 103 of 146  Not logged in ELOG logo
ID Date Author Topicdown Subject Text Attachments
  1011   07 Jul 2014 Ryu SawadaBug Reportmhist does not show history when -s option is usedWhen I use -s option of mhist, it does not
show history, for example.
  
  1021   14 Oct 2014 Konstantin OlchanskiBug ReportProblem in mfe multithread equipmentsIn the ALPHA experiment at CERN I found a
problem in mfe.c handling of multithreaded
equipments. This problem was in 
  
  1022   14 Oct 2014 Konstantin OlchanskiBug ReportHostile network scans against MIDAS RPC portsAt CERN I see a large number of hostile network
scans that seem to be injecting HTTP requests
into the 
  
  1023   14 Oct 2014 Konstantin OlchanskiBug ReportProblem in mfe multithread equipmentsFor my reference:
good version: https://bitbucket.org/tmidas/midas/src/6899b96a4f8177d4af92035cd84aadf5a7cbc875/src/mfe.c?at=develop
first breakage: https://bitbucket.org/tmidas/midas/src/c60259d9a244bdcd296a8c5c6ab0b91de27f9905/src/mfe.c?at=develop
  
  1024   14 Oct 2014 Konstantin OlchanskiBug ReportProblem with EQ_USERIf you use EQ_USER in mfe.c and have multiple
threads writing into the ring buffer, you
will have a big 
  
  1025   14 Oct 2014 Stefan RittBug ReportHostile network scans against MIDAS RPC portsDoing this through the ODB seems ok to me.
If the ODB cannot be accessed, you can fall
back to no protection.
  
  1026   15 Oct 2014 Stefan RittBug ReportProblem in mfe multithread equipmentsYou are absolutely correct, the code is certainly
wrong. It looks to me like the 
  
  1027   15 Oct 2014 Stefan RittBug ReportProblem in mfe multithread equipmentsPlease disregard my previous posting, you
don't need the while loop, since it's already
in the scheduler (around lines 2160 under
  
  1028   15 Oct 2014 Stefan RittBug ReportProblem with EQ_USERSure, each thread needs its own ring buffer
for writing.
  
  1029   16 Oct 2014 Stefan RittBug ReportProblem with EQ_USERI restructured the front-end code to enable
multiple readout threads for EQ_USER equipment.
Last summer I was definitively interrupted
  
  1030   16 Oct 2014 Stefan RittBug ReportProblem in mfe multithread equipments> while (1)
>    wait 10 ms for an event
>    process event, loop back
  
  1031   16 Oct 2014 Konstantin OlchanskiBug ReportHostile network scans against MIDAS RPC ports> Doing this through the ODB seems ok to me.
If the ODB cannot be accessed, you can fall
back to no protection.
  
  1032   16 Oct 2014 Stefan RittBug ReportHostile network scans against MIDAS RPC ports> Sometimes we have very small MIDAS installations,
i.e. just one machine by itself, and such
setups should be secure/secured easily -
  
  1042   27 Jan 2015 Konstantin OlchanskiBug Reportgetaddrinfo()To support IPV6, we need to migrate MIDAS
from gethostbyname() to getaddrinfo(). (Thanks
to 
  
  1072   16 Jul 2015 Thomas LindnerBug Reportjset/ODBSet using true/false for booleansMIDAS does not seem to be consistent (or at
least convenient) with how it
handles booleans in AJAX functions.
  
  1082   29 Jul 2015 Stefan RittBug Reportjset/ODBSet using true/false for booleansSee bitbucket for the solution.

https://bitbucket.org/tmidas/midas/issues/29/jset-odbset-using-true-false-for-booleans#comment-20550474
  
  1094   19 Aug 2015 Pierre GorelBug ReportSequencer limitsWhile I know some of those limits/problems
have been already been reported from
DEAP (and maybe corrected in the last version),
  
  1095   19 Aug 2015 Pierre-Andre AmaudruzBug ReportSequencer limitsThese issues have been addressed by Stefan
during his visit at Triumf last month.
The latest git has those fixes.
  
  1096   19 Aug 2015 Konstantin OlchanskiBug ReportSequencer limits
> See LOOP doc
>  LOOP cnt, 10
  
  1097   20 Aug 2015 Stefan RittBug ReportSequencer limits> > - ODBGet (and ODBSet?) does seem to be
able to take a variable as a path... I
> > was trying to use an array whose index
  
ELOG V3.1.4-2e1708b5