WP8118 DAQ switch upgrade
Dave:
At EY I removed the old netgear switch and mounted the new Brocade switch in its place (last week I had installed the Brocade temporarily, resting on top of the Netgear). This took longer than anticipated due to inaccessible screws severely torqued down.
I have some more data points on what happens to DAQ data when an end station is shutdown. Last week powering off the netgear caused all FECs on h1dc0-card1 to go bad (both end stations, and most of corner ISC). This morning I tried just disconnecting a single front end (h1iscey) from the switch, again this hosed everything on h1dc0-card1. So I gave up trying to nicely and slowly remove FECs and powered down the EY switch, this time only EY went bad, as originally expected. I'm not sure why this behavior is inconsistent.
fmcs-epics-cds upgrade/reboot
Carlos:
The FMCS EPICS IOC machine fmcs-epics-cds was rebooted. The code did not automatically restart. I stopped the alarms, restarted the FMCS IOC and restored the alarms. The IOC was burt restored to 7am.
WP8141 h1sqz model change
Nutsinee, Daniel:
a new h1sqz model was installed. DAQ restart was required.
WP8137 remove PSL-DBB Jitter IPC from h1lsc
Dave:
I installed a new h1lsc model, removing the Dolphin IPC receiver for the PSL-DBB Jitter signal (which has been unavailable since mid-2018 when the Dolphin card was removed from h1psl0). This has greened up the IPC status, allowing me to show it again on the CDS overview. DAQ restart was required.
DMT Nat router upgrade
Carlos, Ryan, Dan:
A new nat router was installed for marble in place of the aging X2200 which was showing a failure LED. The original 10G SFPs did not work in the new hardware, and downgrading the SFP to various types of 1GB was unsuccessful. The original X2200 was put back into place for now. It is no longer showing a failure LED.
Beckhoff upgrades
Daniel, Dave:
Daniel installed new Beckhoff code, resulting in new DAQ-INI files for C1PLC[2,4], Y1PLC[1,2], X1PLC2. DAQ restart was required.
Some of the Beckhoff SDF faux-models were unresponsive, I restarted them all on h1ecatmon0. For each new channel which was not initialized, I (ACCEPT+MONITOR)'ed them.
Not Done This Week:
Ran out of time due to unexpected difficulties meant that external EDCU, DMT cable cleanup and EX switch upgrade will be deferred to next maintenance or TOO.
h1lsc on overview no longer has its IPC notched out.
Restart log for this morning's activities. Note that beckhoff PLC restarts are not logged here. All times PDT, we overran end-of-maintenance by 35 mins.