Jonathan, Tony, Erik,
As per WP 13402 the h1daq*1 systems where updated today. We removed the Dolphin PX cards from the systems, installed the new ethernet cards.
In the prep stages the GDS broadcast was moved to the 0 leg and h1daqfw2 was changed to pull data from the 0 leg.
Notable changes:
* h1daqgds1 -> is now the the data concentrator + broadcaster. It is named h1daqdc1. This is the machine that receives the front end data.
* h1daqdc1 -> is now h1daqkc1 (h1daq kafka connector) for use with the ngdd project.
We found one issue where the daqd could not write jobs into the nds servers job queue folder. This was a permissions issue. Not sure why this popped up as a problem now. I updated the puppet config specifically avoid this happening again.
To help with load testing, h1daqfw2 was moved to reading data from dc1, and kc1 is pulling data from dc1 even though it is not publishing data yet (only kc0 does at this point).