USNO VLBA Support Meeting
*February 14th, 2013 Time 11 am Socorro, 1pm Eastern *
Phone: 575-835-7170
USNO Attendees: Dave Boboltz, Alan Fey, Nicole Geiger, Chris Dieck, John Spitzak
NRAO Attendees: Steve Durand, Peggy Perley, Walter Brisken, Ephraim Ford, James Robnett
Agenda
- Previous Action items
- NRAO: January Weekend Outage Actions -
- NRAO (James) Will revise procedure to notify outage properly.
- Had an actual outage (Feb 5th) due to new network switch. The outage was short, a few hours, but throughput was reduced for the next 6 days.
- NRAO (Walter) Revise script to use different baseline. Done.
- New mode in script called GOUSNO. Operators can exercise this. If you specify 2+ antennas it will drop out of mode using USNO Mk5c & DBE and use NRAO units. This was tested on MK & LA. In actual usage we would rush delivery of modules, hopefully giving arrival within 3 days. NRAO would correlate.
- Scenario definitions needed: If PT is down, which base line? If MK is down, which baseline? If both, which sites? It would be good to have multiple options in case of media shortage.
- Action USNO: Dave to provide Walter options to all potential backup baseline scenarios above.
- If link is down, data will be recorded and sent to Socorro for correlation, then to USNO.
- NRAO (Walter) will define false failure exercise to check new procedures above.
- USNO: St. Croix as backup if other site is down. Dave to discuss with Steve if this is feasible or interesting
- tabled for future discussion.
- Current rough cost estimate is based on a lower fiber speed. Mothly cross-SC fiber cost is $5k/month. Then to Jacksonville and on.
- NRAO will sharpen estimate of St. Croix usage. USNO will provide Jacksonville contact.
- Review of Monthly Summary of Daily Observations, January - Peggy
- 3 days of issues in January (5-7). 3 consecutive days in MK and latter two at PT.
- There were 2 days with a warm RX in MK, data was usable by USNO.
- End of month saw no vex.preobs delivered. Correlated fine so USNO didn't notify NRAO. Will do in future.
- Update from Walter: It was not a software change, but an OS upgrade on one machine that broke a library dependency that prevented db2vex from running properly. Operations has since switched to using a machine that has the right version of everything.
- Review of Quarterly Report - Steve
- Will submit bill on previous PT-MK support contract (contract 3, not 4).
- VLA Network change and its impact on PT data transfer rate - James
- Had an actual outage (Feb 5th) due to new network switch. The outage was short, a few hours, but throughput was reduced for the next 6 days.
- Some packet loss is still seen, possibly a bad fiber in network.
- Due to local USNO bottlenecks, they have not tried transfer above 150 MBps but have not seen problems at that rate. James doubts full throughput (4-500) would work
- Action USNO: Throttle up to full throughput to see whether it is achievable at USNO despite NRAO-side issues. James would like to see results.
- Open issues
- none, see below for other discussions.
Discussion Minutes
- NRAO:
- ~10 days ago NRAO placed a fix for problem of same file names showing up. Please advise if seen.
- Fix also hoped to solve missed scans but doesn't seem to. Keep those reports coming. Possible fix for this is to increase scans by a few, increasing observation by a few minutes.
- Walter will explore how much extra observation time he can slip in without affecting observations.
- Maser at MK had two ion pump issues. NRAO is investigating the leasing of a laser for replacement and closely tracking performance.
- USNO:
- Expects to be caught up within 1-2 months and able to do same or next-day analysis of data.
--
EphraimFord - 2013-02-11