NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Software > PlanOfRecordC62007

Changes | Index | Contents | Search | Statistics | Go

Software Plan of Record

Cycle 6 2007 (August 16 - September 30, 2007)

• 1. Operational/User Support Items Completed During C6 2007
• 2. Commitments
• 3. CM&E
• 4. Background Tasks

1. Operational/User Support Items Completed During C6 2007

For each support item, please list the amount of effort involved as either: low, medium, or high. Guidelines for these rankings are as follows:

• low - Things that take less than two hours.
• medium - Things that take more than two hours, but less than eight hours.
• high - Things that take more than eight hours.

If multiple people are involved in an item, please list it only once. Effort estimates should reflect total time elapsed, i.e. how long did it take to fix the problem start to finish - not the sum total of each person's time.

• date - name - effort
• 17 Aug 2007 - JoeBrandt - Recover from multiple power failures - medium
• 24 Aug 2007 - RamonCreager - Patched correct Spigot sources and executables into 7.5. low
• 31 Aug 2007 - MarkClark - Antenna failing to update scan log entry, fixed by restart. low
• 05 Sep 2007 - MarkClark - RcvrPF_1 manager crashes because MCB not connected. low
• 07 Sep 2007 - PaulMarganian, JoeBrandt - GFM cannot plot peaks due to NULLs in Antenna data. low
• 11 Sep 2007 - MarkClark, JoeBrandt, RamonCreager - Rcvr12_18 manager crashes because MCB not connected. low
• 13 Sep 2007 - MarkClark - MMConverter manager crashes because MCB not connected. low
• 03 Sep 2007 - 21 Sep 2007 - MarkClark, RonGrider - LPCs fail to transfer from gfm to turtle (ongoing). medium
• 18 Sep 2007 - BobGarwood - ToneyMinter reported GBTIDL hanging when applying flags. Fixed in gbtidl-test, see CM&E item. low
• 19 Sep 2007 - MarkClark - Ka receiver failed to talk to MCB until MCB was reset. low
• 21 Sep 2007 - RamonCreager, MarkClark - MCB communications failed on wind. Restarting MCBServer did not help. Though MCBServer did not report any problems opening or configuring /dev/ttyS0, it nonetheless could not communicate using the serial port (mcbcmd's 'siblist' command returned immediately, for example, whereas it should spend some time probing for every possible SIB address). Rebooting wind solved the problem. low
• 22 Sep 2007 - PaulMarganian - Observer could not connect to data in GBTIDL in 'online' mode; turns out online sdfits did not get restarted when datacapture was rebooted earlier the previous day. low

2. Commitments

The default/unsorted state reflects priority order for this cycle.

Project	Task	Info/MR	Lead	Sponsor,Team	Notes	Status
E2E	Create SRS for PST Administrative Functionality	create Word document	MikeMcCarty	NicoleRadziwill, AmyShelton, CarlBignell	Once Carl has finished with other tasks, he will take the lead in writing this document. He should be free in a couple of weeks.
Data Reduction	Evaluate GBT Data Handling Status & Future Options	Work in Progress	BobGarwood	AmyShelton	On hold while the other MRs have my attention.
KaBandReceiver	Control of Subreflector Nodding by Astrid	ModificationRequest2C407	MarkClark	DjPisano, MikeMcCarty
KaBandReceiver	Subreflector Nodding Observing API Building Block	ModificationRequest17C507	MarkClark	DjPisano, MikeMcCarty	Complete except for testing issue solution which needs to be approved by the sponsor.
KaBandReceiver	Indicate Subreflector Position in Data	ModificationRequest4C507	BobGarwood	RonMaddalena, MarkClark, MikeMcCarty	Written except for one possible item that would be useful to test during if possible. The checking step can begin now.
KaBandReceiver	Implement GBTIDL Reduction Procedure for Subreflector Nodding	ModificationRequest5C507	BobGarwood	DjPisano, RonMaddalena	Another round of feedback send back to Ron and others.
KaBandReceiver	Proper Setting of LO1 B Frequency To Avoid Interference	ModificationRequest10C107	MelindaMello	GalenWatts, RogerNorrod
KaBandReceiver	Finalize Ka FITS File Specification	ProjectNotes	KarenONeil	KarenONeil, MarkClark, MikeMcCarty
KaBandReceiver	Ka Manager Changes	ModificationRequest1C507	MarkClark	GalenWatts, MikeMcCarty	No change, GalenWatts is working on MR.
KaBandReceiver	config_tool Changes Related to Ka Manager Changes	ModificationRequest7C507	MelindaMello	GalenWatts
KaBandReceiver	IF Manager Changes Related to Ka Manager Changes	ModificationRequest8C507	MarkClark	GalenWatts, MikeMcCarty	Not started, need new Ka manager MR (ModificationRequest1C507) first.
KaBandReceiver	config_tool Changes Regarding LO1B/mm-wave Converters	ModificationRequest10C507	MelindaMello	GalenWatts, RogerNorrod
KaBandReceiver	Correct Labeling of Ka-band Hybrid-Switched Signals	ModificationRequest3C407	MarkClark	DjPisano, MikeMcCarty	MR complete, but tabled.
KaBandReceiver	Fix GFM's Processing of Ka Data	ModificationRequest9C507	MarkClark	RonMaddalena, MikeMcCarty	Tabled.
KaBandReceiver	Merge Zpectrometer code branch into main trunk	ModificationRequest15C507	BobGarwood	KarenONeil, AndrewBaker	Done except for testing zpectrometer online filler and data access.
KaBandReceiver	Update AIPS++ Filler to Handle Ka Changes	ModificationRequest1C607?	BobGarwood	KarenONeil
PTCS	General PTCS Support	N/A	RichardPrestage, JohnFord	JoeBrandt, RonGrider
PTCS	Implement PFM5 revised pointing/focus model	ModificationRequest1C407	JoeBrandt	ToddHunter	Added Model5a and Model5b into PFM database.
PTCS	Port Holography Manager to Linux	ModificationRequest1C307	RonGrider	JohnFord, JoeBrandt	Tested with sponsor, with Holography backend hardware. MR signed 20 Sept, awaiting confirmation of integration test results from StevenWhite.
PTCS	Add DIO interface to Linux Holography Manager	ModificationRequest12C507	RonGrider	JoeBrandt, ToddHunter, JohnFord	(No change) Revising MR to reflect recent team discussion about interrupt characteristics of new interface card, and the effects this will have on software design.
PTCS	Complete Pointing Processing Description Memo	Should become PTCS/PN/23 - Add link to document	JoeBrandt	RichardPrestage	Needs some cleanup to include focus tracking.
PTCS	Modify Peak Finding Algorithm for Pointing/Focus Code	ModificationRequest4C207	RonGrider	FrankGhigo, JoeBrandt	FrankGhigo and ToddHunter have settled on GaussianFilter as preferred method. However, since this approach changes the resultant pointing offset values, deployment of the selected approach will be postponed until after the new pointing model is implemented this summer (ModificationRequest1C407). Deployment on HOLD.
PTCS	Analysis of Trajectory Generation Problems	add link to analysis	JoeBrandt	JohnFord, RonGrider
PTCS	Find/Fix Command Stream Glitches	add link to analysis or eventual MR	JoeBrandt	JohnFord, RonGrider
PTCS	SCCU Servo Changes	ModificationRequest1C806	JoeBrandt	JohnFord	Although the code is working, we have found that commanded velocity spikes occur during normal observing. To date this has not been a problem since the hardware filters out the spikes. However with the hardware modifications coupled with this MR, the spikes will corrupt the intended operation of the friction compensation. John is working on modeling a different methods of deriving the velocity commands.
PTCS	Create Pyrgeometer Manager	ModificationRequest2C207	JoeBrandt	JohnFord, RonGrider
PennArrayReceiver	General Mustang Support	N/A	MelindaMello	BrianMason, RamonCreager, PaulMarganian
PennArrayReceiver	Mustang Meta-MR	MasterMustangManagerMr	MelindaMello	BrianMason, RamonCreager, PaulMarganian	RamonCreager - Added timestamps to PCI card driver. Testing in progress. Main.PaulMarganian - Crate and Tower MR work complete at this stage, moving on to FITS Writer (see below). Group discussion s started on design of data buffering between data stream and data consumers.
PennArrayReceiver	Data File Stream Upgrade	ModificationRequest1XC507?	MelindaMello	BrianMason, RamonCreager, PaulMarganian
PennArrayReceiver	Mustang Crate Control	ModificationRequest16C507	MelindaMello	BrianMason, RamonCreager, PaulMarganian	Brian has approved MR. Ygor parameters have been refactored and tested w/ hardware, all successfully, except for the exceptions noted in test section of MR. I anticipate no further work on this will be needed until other parts of the system are integrated, and/or sponsor testing.
PennArrayReceiver	Mustang Tower Control	ModificationRequest14C507	MelindaMello	BrianMason, RamonCreager, PaulMarganian	MR approved by sponsor. Tower Ygor Parameter's refactored successfully, but of course, no hardware to test with. No further work on this anticipated in the near future.
PennArrayReceiver	Mustang PCI Initialization	ModificationRequest3C607	RamonCreager	BrianMason, MelindaMello, PaulMarganian	Under sponsor review.
Cicada	GPU R&D	Presentation1, Presentation2	JoeBrandt	AmyShelton	Two nodes ready for Wolfgang to install. Created status page. Been looking into clustering tools/environments.
Dynamic Scheduling	Dynamic Scheduling Software	Dynamic Web	AmyShelton	KarenONeil, MelindaMello, PaulMarganian, MarkClark, MikeMcCarty	No progress.
BusinessOfficeSystem	Business Office System	BusinessOfficeSystem	AmyShelton	AmyShelton, MikeMcCarty, PaulMarganian	Release version 0.5 for sponsor testing. See To Do list for items fixed for deployment (mostly done) and items for future releases. 2 of 3 MRs sponsor tested.
BusinessOfficeSystem	Meta MR linking other MRs and other issues	ModificationRequest9C107	AmyShelton	MikeMcCarty, RamonCreager, PaulMarganian	The deployment configuration is setup; include SSL! Some changes still need to be made to the code that accommodate HTTPS redirects.
BusinessOfficeSystem	User Management	ModificationRequest5C107	AmyShelton	AmyShelton, MikeMcCarty, RamonCreager, PaulMarganian	Complete.
BusinessOfficeSystem	Database Migration	ModificationRequest6C107	AmyShelton	AmyShelton, MikeMcCarty, RamonCreager, PaulMarganian	Updated MR text and send to sponsors for review. Working on BARS schema using SQLAlchemy.
BusinessOfficeSystem	Reservation Management	ModificationRequest7C107	AmyShelton	AmyShelton, MikeMcCarty, RamonCreager, PaulMarganian	Complete.
BusinessOfficeSystem	Document Management	ModificationRequest8C107	AmyShelton	AmyShelton, MikeMcCarty, RamonCreager, PaulMarganian	Complete.
Software CM&E	Integration Testing	SoftwareReportCentral	AmyShelton	MarkClark
Continuing Education	TurboGears Tutorial	Tutorial Materials	AmyShelton	MikeMcCarty, RamonCreager, PaulMarganian	Done.
PTCS	CCU Servo Interlock Update	ModificationRequest03C042007	JoeBrandt	JohnFord, TimWeadon	Now have mockup CCU/CCU-aux running with code changes for new servo interlocks setup in the basement. Still having problems with getting CCU software to run properly without hardware. Sent MR to sponsor for approval.
PennArrayReceiver	FITS Writing	ModificationRequest18C507	MelindaMello	BrianMason, RamonCreager, PaulMarganian	Just started this today - gathering information on how other Backends do this, and is part of wider discussion concerning data buffers and expected data rates.

3. CM&E

Description	Tests	Submitter	Notes	Status
Subreflector tracking problems/data dropouts SubreflectorTracking	With the antenna setup for Gregorian mode, start a sampler2log Antenna AntennaManager? subreflectorAct and examine data for drops (Seen as jumps in timestamp values.)	JoeBrandt	Code ready for patching.
MCB hardware failure causes Manager coredump on startup	Start a manager that has an MCB component that is missing/down (for example, a receiver manager whose receiver is down for maintenance). The manager should start and remain running. A message should appear in its log file stating that it could not communicate with its SIB. Messages to this effect may also appear. If this is the case, they should clear if the manager is turned off.	RamonCreager	The SIB class was not constructing if the SIB hardware was not present because of an uncaught exception in the SIB constructor. This may potentially leave an invalid pointer in a manager's constructor if that manager catches the exception. If this pointer is later used, the manager segfaults. The SIB class now catches the exception, allowing construction to succeed even if the SIB hardware is absent, preventing the segfault. The SIB class already includes a mechanism to recover when the hardware comes back online.
Applying flags in GBTIDL can trigger an infinite loop	Flag several channel ranges in one scan with several integrations. Make sure that there are more channel ranges than there are integrations in that scan. When you next try and get that data IDL will hang when this bug is present.	BobGarwood	The inner loop processing the set of channel ranges re-used the same loop variable used in stepping through each integration on the outer loop. It was a simple fix in the end. This fix is now in gbtidl-test.
Added a 'repeat' parameter to the Zpectrometer manager.	Set the 'Simulate' flag to 'true' in Zpectrometer.conf and restart the manager. Select an 'obsProgram' that uses the 'repeat' argument such as 'zero'. Run a scan and look at the manager's stdout output. Verify that 'zero' was run with the command-line argument 'repeat=1'. Now, using Device Explorer or some other means, change the 'repeat' parameter to another value, such as '2'. Repeat the test and verify that 'zero' was run with the command-line argument 'repeat=2'. Set the 'Simulate' flag back to 'false' and restart the manager.	RamonCreager	This capability will enable the Zpectrometer to use the secondary nodding observation modes.

4. Background Tasks