| MR | Developer | Release | Sponsor Tested | Integration Test Time (HH:MM) | Maintenance Test Time (HH:MM) | Comments |
|---|
| Astrid Hangups | MarkClark, JoeBrandt |  | | 4:00 | 4:00 | A short period of heavy peaks and focuses, the remainder testing can piggy back on other observing. |
| GFM/Turtle Upgrade - Remove use of LPCs.conf file | RamonCreager | | | | | Can be tested concurrently with Astrid Hangups above. |
| Modify SDFITS and aips++ to correctly process Spectrometer cross-polarization data | BobGarwood | | | | | Testing has been done offline. |
| SDFits Changes - Add target position and CALTYPE keywords to sdfits and other requested sdfits changes | BobGarwood | | | :00 | :30 | A frequency switched scan should check the Tsys calculation in GFM and if the coordinates are RA,DEC then that checks other things. A few scans in a row could be used to check the online filler with gbtidl-test to make sure gbtidl-test can read the index without wanting to remake it and that it gets the new information correctly (the target position, caltype, and radesys). Finally, a scan using the high cal should be taken. |
| GBTIDL Changes necessary in support of related sdfits tool changes. | BobGarwood | | | | | same tests to be covered by MR 3C706 |
| Change scaling for the Autocorrelation Spectra in the Spectrometer | BobGarwood | | | | | Testing has been done offline. |
| Change GO fits file so that RA is in degrees | PaulMarganian | | | :15 | :15 | After a scan, insure that the GO FITS file RA value is in degrees, and not hours. This could probably be done in parallel with other tests. |
| Upgrade PTCS Pointing Model Pre-Processing Software | RonGrider | | | | | Testing this application can be done 'offline'. Refactored code to support this application requires testing time, per section entries below. |
| Get, Set, and Change attenuators. | MelindaMello | | | | | |
| Change Spectrometer Warning Messages During Balance | MelindaMello | | | | | |
| IF Rack: Change messages to indicate non-balance and non-linearity. | MelindaMello | | | | | |
| Multiple sampling for Balancing procedures to diminish RFI effects. | MelindaMello | | | | | |
| SCCU Servo Changes | JoeBrandt | | | | | |
| Update GFM Tsys Calculation for the Spectral Line Plugin | BobGarwood | | | | | Verify that Tsys values in GFM are close (<1%) to the values report by GBTIDL on the same data. </td> |
| Extract Sampler Information from Log Files & Create Antenna Characterization file | JoeBrandt | | | | | |
| Config Tool: Modify Default IF Target Levels | MelindaMello | | | 0:45 | 3.0 | This includes 1 hour for sponsor testing. Sponsor testing can not occur 1-24 to 1-26 and on 1-29 from 10:2 and on 2-1 from 9-12. Jim can run his sponsors tests on this if he wants during Regression Testing |
| Configuration Tool Default Value Changes | MelindaMello | | | 0:30 | 1.5 | This includes .5 hour for sponsor testing. Sponsor testing can not occur 1-24 to 1-26 and on 1-29 from 10:2 and on 2-1 from 9-12. Jim can run his sponsors tests on this if he wants during Regression Testing |
| Description | Developer | Release | Integration Test Time (HH:MM) | Maintenance Test Time (HH:MM) | Test Procedure |
|---|
| shm_ntp_clk | JoeBrandt | | | | shm_ntp_clk is the ntp clock helper program used on systems with a 1PPS input. No specific tests necessary. |
| LO1Coordinator | JoeBrandt | | | 0:15 | Select PAR receiver, turn auto-leveling on. Check for message. |
| sampler2log | JoeBrandt | | | BG | run sampler2log -d 0.09 Antenna AntennaManager azElData and verify row timestamps are spaced approximately 100ms apart. |
| Added check for valid beamName to ScanTypes | AmyShelton | | | | No additional testing needed. Unit tests pass. |
| Refactored *Device classes | AmyShelton | | 0:20 | | Run GFM and exercise each plugin to make sure you get plots. The "GFM Workout" SB in TAPI will exercise the peak, focus, and spectral line plugins. |
| Removed use of placeholder BackendDevice from GFMPanel | AmyShelton | | 0:15 | | Run GFM and exercise the Spectral Line plugin. It should work the same as before. |
| Added basic session checking to Run tab of turtle client | AmyShelton | | | 0:05 | For projects whose names begin with "A", users should only be able to specify integer session values. If a non-integer is used, the session field turns red and the user should not be able to submit SBs. This can be done in parallel with other testing. |
| Modified Grail to provide full field callback support | RamonCreager | | 0:15 | | Client should be able to register field callbacks for simple type samplers and parameters (e.g. <manager>.reg_param('state', cb, '.')). Clients should also be able to obtain the time stamp MJD and seconds values by using the 'TS:MJD' and 'TS:seconds' fields, for both samplers and parameters. |
| ScanCoordinator core dump | JoeBrandt | | | 0:25 | Fixed an issue where a string parameter of fixed size (scanLogEntry) was being copied from a smaller buffer. This caused a seg fault when the unmapped memory was accessed. |
| TaskMaster core files | JoeBrandt | | | 0:15 | Fixed Taskmaster to properly place core files into the cores directory. |
| Misnamed FITS files | JoeBrandt | | | 0:30 | Fixed the algorithm for computing device FITS filenames. Problem was seen with a starttime of 16:20:00.0 which was incorrectly computed as 16:19:00. |
| Complete Logview and Deploy | MarkClark | | 0:00 | 0:00 | LogviewTests |
| JIRA GBT-37 | BobGarwood | | | | The factor of 0.99 is now being divided instead of multiplied. GBTIDL unit tests updated. |
| JIRA GBT-38 | BobGarwood | | | | GBTIDL unit tests still pass, new unit test for trackfdnum=1 added for getnod and getbs. |
| JIRA GBT-22 | PaulMarganian | | 0 | 0 | /home/sdfits has less then a months worth of log files there, more then a month since this fix was patched in. No online tests necessary |
| JIRA GBT-15 | PaulMarganian | | 0 | 0 | GFM behaves sanely in simulator now. Should be a good enough test, since this behaviour has never been seen with production system. We have no known way of testing this outside the simulator, but problem only occurs in simulator. No online tests possible. |
| JIRA GBT-7 | PaulMarganian | | :15 | :15 | unit tests contain the original data set that illustrated the problem. Also created several new unit tests to make sure that sdfits and GFM center frequencies agree.
Bug is fixed, but none of the refactoring or documentation recommendations in JIRA entry have been started. Online test: make sure spectra agree between gbtidl and gfm. |
| JIRA GBT-39 | PaulMarganian | | BG | BG | full test should wait for integration/regression tests, where all systems are running. First tests can also happen on maintanence day. Test basically makes sure that all processes are running. Note: this requires switching of *.proc files during versions switch. |
| astrid console import problem | MelindaMello | | 0:05 | 0:05 | click the "console" tab in the lower frame. A message should appear in the frame saying that the configuration and balance apis were loaded sucessfully. |
| JIRA GBT-16 | PaulMarganian |  | | | * lots of new unit tests exist in sparrow/api for supporting IF Numbers
* need explicit tests to compare sdfits and gfm IF number results. |
| JIRA GBT-42 | PaulMarganian | | :05 | :05 | after version switch, make sure these aren't running on earth and wind via 'top' and 'Task Master'. Make sure to tell the operators! |
| Add VLBA_DAR as legal pulsar backend | Main.Melinda | | | 0:25 | Mark and I arranged for Frank to test this at 8:00 on the first maintenance day we have time for during the week beginning 1-22-07 |
| Refactored DEAP Config file | RonGrider | | 00:05 | 00:05 | Jim Make changes to sparrow.conf file, open GFM and check that corresponding setting has changed. Concurrent with refactored pointing / focus scan below. |
| Refactored Pointing and Focus scan | RonGrider | | 00:30 | 00:30 | Jim Exercise Peak and Focus scans in GFM, make sure plots and LPC / LFC calcs appear to behave as before. |
| Mean error stats in fitting methods | RonGrider | | 00:30 | | Need to exercise scans that would be fitted by all fitting routines. Make sure GFM appears to operate as before, that nothing is broken. Offline analysis (provided by RG) to check data captured while exercising Peak and Focus scans in GFM above. |
| Changed ifpath pickler to use new ports for the 26-40 receiver | MelindaMello | | | 0:15 | I have arranged for Ron to sponsor test this on Thursday afternoon (sometime) on thursday 1-25-07. IMPORTANT NOTE: This requires a change to the cabling file. I have asked Galen Watts to generate a cabling file for Thursdays testing and will let the SDDEV know of its name and location as soon as I hear from Galen. |
| Changed config tool to send an LO1 manager activate after configurations using the CCB | Main.Melinda | | | 0:15 | |
| Added Zpectrometer panel to Astrid | AmyShelton | | | 0:05 | In Astrid, select the File menu and then select Zpectrometer from the list of candidate windows to open. The Zpectrometer panel should appear as a tab and be populated with various sampler values. These tests are passive and can be done in parallel. |
| Added Quadrant Detector panel to Astrid | AmyShelton | | | 0:05 | In Astrid, select the File menu and then select Quadrant Detector from the list of candidate windows to open. The Quadrant Detector panel should appear as a tab and be populated with various sampler values as well as updating plots. These tests are passive and can be done in parallel. |
| JIRA GBT-43 | BobGarwood | | | | Make sure each invocation of sdfits on the same data results in an index file of the same size as before. Only -append should lead to a larger index file. No online tests possible. |
| JIRA GBT-51 | BobGarwood | | | | Specify a tcal value to any GBTIDL calibration routine and verify that !g.s[0].mean_tcal is that value after the calibration has finished. |
| SourceForge GBTIDL #1644760 | BobGarwood | | | | Suppose scan 6 has procedure "Track" and scan 7 and 8 have procedures "OnOff". select, scan=[6,7,8], procedure="OnOff" should succeed where previously it would fail with a cascade of error messages. |
| JIRA GBT-45 | PaulMarganian | | | | Make sure that 'dirin, /new_index recreates all index files. No online tests possible |
| GBTIDL improved default ap_eff values | JimBraatz | | | | Verify that new values are being used. |