Milestone Nov 1, 2019–May 8, 2020
Primarily targeting Debian 10 Buster.
This should be ready for deployment on the LIGO production systems once we come out of quarantine.
Unstarted Issues (open and unassigned)
Ongoing Issues (open and assigned)
Completed Issues (closed)
- RCG no longer producing caQtDM screens on front-end builds
- DAQ KB/S rates do not match between old daqd and producer
- Add EPICS outputs to stand-alone EDCU
- EzCaRead.pm has mismatch in Conn/Err output
- The standalone_edc doesnt export the epics variables it claims to
- Restrict model start in Puppet installs to right host
- High dcu (ecat) models added to par
- The standalone edcu shouldn't need a .par file for local_dc to work with it.
- The systemd unit for the standalone edc is incorrect, uses 'I' instead of 'l'
- rtcds build World stops at first error
- standalone_edc overflows the stack
- IOPs running on secondary timing in test stands aren't updating.
- When a GDS broadcast list is specified and it contains channels that do not exist the daqd should crash
- Compile error building some models
- Errors in awg output when run from systemd
- Update 'rtcds lsmod' script to support Dolphin IX, replacement for open-mx
- Convert the advligorts-transport-dix to use the new ligo-dolphin packages
- cps_xmit does not need the -d gds_tp_dir param.
- The receiver in the pub_sub code does not properly detect late messages
- cdsrfmmonitor.py calls python2 scripts
- Specialized install scripts not installed with RCG
- cdsrfmswitch.c not installed
- Ensure that non-real-time models are built and started correctly
- Archival data retrieval not working from rts-nds service
- Add cds_pub_sub to the advligorts-transport-pub_sub package
- dolphin-ix package install failing on non-realtime systems
- DAQ GDS broadcasters need to use different frameXmitPorts
- Need method to restart data concentrator, downstream daqds to synchronize channel list changes
- awgtpman cannot set nice when run as non-root user.
- update default daqdrc entry thread_stack_size
- Should remove athena sub-directory or migrate to its own repo
- Downstream daqd stop when front-end rebooted causing data concentrator DAQ restart
- read buffer too small for large IFO daqd
- ADC auto-calibration status not checked
- run awgtpman as non-priviledged user
- Clearing DAC watchdogs on SUS machine without I/O chassis causes lockup
- local_dc returning zero exit code failure
- 'setup' config for particular deployments should be moved to a separate repo
- The advligorts-daqd sample daqdrc is installed in such a way that by default daqd will not start.
- We should add RT scheduling capability support for daqd in the systemd service files (and maybe the transports too).
- Add support for module versioning to pcie_switch code
- SWSTAT epics p.v. missing ramp bits from SW2R
- Add support for IPMI to Debian 10 RT config
- Add support for module versioning of Dolphin kernel modules
- Add some signaling via EPICS as to when the raw trend files are being written
- find_module_symvers failing
- new bootstrap breaks RCG built procedure
- Remove old etc/install scripts
- Commit Debian 10 RT kernel config file
- Fix feCodeGen.pl for timing slaves
- support heterogeneous DC receivers
- remove use of burt from RCG
- The streamers and daqd_shmem need to have a higher dcuid limit.
- CRC errors between front ends and DAQD with local_dc and omx_xmit as transport
- unbreak restart procedure for dual data concentrators
- using EXPORT_GPL'ed symbols in a module that is unlicensed/doesn't declare a license
- modvers/symvers not being generated/used correctly
- The daqd/nds1 server has a very low connection limit
- long IOP cycle times during startup of front end in deb 10
- cmake doesn't recognize libfl-dev requirement
- remove '_shmem' suffix from daqd binary
- unnecessary linking in nds, daqd_shmem, dv