MPC Status Page: Archive (2004 August)

Recent problems are listed elsewhere. Index of other older problems..

Older Enhancements and Resolved Problems

• MPEC 2004-Q57

  2004 August 28: 12:00. It seems there was a production problem with last night's DOU MPEC that prevented the new NEO observations from being included. We are investigating, the observations should appear on tonight's circular.

• More "Security" Edicts from the Computation Facility

  2004 August 26: 15:40. We have received notification from the Computation Facility that all passwords to their CF boxes must be changed before Sept. 7 and that they must now be changed every 90 days. The MPC has numerous command files that copy material over to the CF webserver. Most of these are executed by entirely automatic processes that run at intervals varying from minutes to months. Some command files are generated dynamically, others are static. The dynamically-generated command files are not affected by this edict. The static files will have to be modified to be dynamically generated each time the relevant processes are run. This is a major headache and the changeover will probably take several days to accomplish. It is possible that some command files will not get changed (or will get changed incorrectly--it isn't always possible to test run a modified command file), so some pages may not update properly after September 7.

  This author's opinion is that this change will actually reduce the level of security on the CF boxes, as users may now have to begin writing their passwords down instead of committing them only to memory. That is what I'm going to have to do...

• Minor Planet Checker

  2004 August 19: 12:00. It seems that a nightly procedure to update the element files used by the Minor Planet Checker has been silently failing possibly for as long as a month. This means that newly-generated orbits are not being included in the search. The missing elements are being reinserted into the element files.

• New observer options in MPES and EOSJT

  2004 August 15: 23:30. A number of new observer options have been added to the MPES and the EOSJT: suppression of ephemeris data when sun is above local horizon; suppression of ephemeris data when object is below local horizon; and allowance for quick selection of non-0h-UT daily ephemerides.

• Patch installation on cluster

  2004 August 13: 10:30. Today we are installing a number of patches on each of the cluster machines, in order to clear up a number of minor problems (none security related...) with the latest version of the OS. There will be brief outages (hopefully under 30 mins a time) of the webserver (and other machines on which it relies) throughout the day.
  • 12:47. The last of the machines was patched at 12:38, ahead of schedule.

• Perturbed NEA elements for non-standard epochs

  2004 August 7: 12:00. A user reported that a known NEA was 30' off the prediction returned by MPChecker. An investigation showed that the routine that calculates elements for non-standard epochs was occasionally outputting incorrect values for certain types of orbit when requesting element output at widely-separated intervals. This has now been fixed. Since MPChecker uses elements at the current 40-day date generated by this program, this problem manifested itself by degrading MPChecker's output. The standard-epoch orbits from last night's DOU MPEC are being rerun through the corrected program. The files used by MPChecker will get rebuilt as part of this procedure.
  • August 8: 12:00. NEA orbits published on MPECs since the last full MPC batch are being rerun through the corrected code to fix the problem. Just to make it clear, the orbits as published on the circulars are correct, the problem lay in elements prepared for certain non-standard epochs.

Index to the CBAT/MPC/ICQ pages. Credits

MPC homepage