MPC Status Page: Archive (2004 March)

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

Older Enhancements and Resolved Problems

• MPChecker

  2004 Mar. 24: 10:57. For objects with perturbed orbital elements, MPChecker now uses elements at the nearest 40-day epoch date rather than at the standard 200-day epoch date. This will help improve the accuracy of the "quick look" positions, especially for NEOs.

• Mid-month-MPS-complete batches of observations in ECS

  2004 Mar. 23: 17:00. A user reported anolmalously large sizes for for the complete-through-the-latest-MPS-batch files of minor planet observations. This problem is related to the March 21 diskspace problem and was exacerbated by the non-logical functioning of a Unix utility. The files are being regenerated.
  • 18:52. The unnumbered file has been regenerated.
  • 21:47. The numbered file has been regenerated.

• Planned radar observations at Goldstone

  2004 Mar. 22: 14:50. The MPES now flags those objects that are scheduled (or have been requested) for observation by radar from Goldstone. Comments are made as to whether astrometry or physical studies of the objects have been requested.

• Display MPEC script

  2004 Mar. 22: 12:15. A user reported a problem with the form at the bottom of the Recent MPECs page, specifically that one could not access circulars numbered above 99. The script has been updated to fix this problem.

• MPEC orbits in MPES

  2004 Mar. 22: 11:00. A user reported a problem with the orbit of 2004 FY3 downloaded from the MPES shortly after the discovery MPEC was issued. Instead of the perturbed orbit used on the MPEC, an unperturbed orbit was returned. This turns out to be a problem with the MPEC production code and it would only appear if a particular sequence of orbit calculations occurred for the object being MPECed. The problem code has been located and tightened up to prevent this happening in the future.

• March 21 DOU MPEC and Mid-Month MPS batch

  2004 Mar. 21: 10:10. One of the cluster-shared disks filled up late last night. Remedial action (purging the disk) was not undertaken at that time. As a result the DOU MPEC failed to run. It has been abandoned.

  The problem also affected the preparation of some files associated with the mid-month MPS batch. The affected files are being regenerated.

  • Regenerated files are now available.

• Erroneous line on today's MPEC

  2004 Mar. 19: 11:57. The four MPECs issued today have an erroneous line after the observer details on the copies that were mailed out. The MPEC production has been modified to prevent these lines appearing in the future.

• 2004 FH orbit on DOU MPEC

  2004 Mar. 18: 10:40. The orbit for 2004 FH published on the latest DOU MPEC had an internal flag set incorrectly. This caused the problems reported by some users. A new orbit appears on MPEC 2004-F26 that corrects the problem.
  • 16:06. A user reported a problem with generating a 1h-interval ephemeris starting on March 16. This appears to be a problem with an internal pointer getting confused by the use of a non-standard epoch for the 2004 FH orbit. This may take time to fix as it will require examination and modification of a library routine. Use of the MPES for current ephemerides is not affected by this problem.
    • 17:12. A fixed version of the code, linked to the modified offending library routine (one of those WTFIsThatDoingThere? moments, in a bit of code called perhaps once in a million runs!), is now on-line.

• Special-Epoch NEA element files

  2004 Mar. 17: 10:40. A user reported that the special-epoch NEA element files on the ftp site contained only one set of elements. The reasons for this are unclear. The files will have to be regenerated from internal copies.
  • 12:20. After making some changes to internal procedures, regenerated versions are now available.

• MPChecker files

  2004 Mar. 15: 10:00. A user reported that MPChecker was not identifying objects. A check showed that the files used by MPChecker had been truncated. The affected files are being rebuilt.
  • 11:12. The affected files have been restored.

• Residual blocks in the MPES

  2004 Mar. 12: 18:04. A user reported that residual blocks were missing from the MPES. A check showed that the procedure that copies the residual blocks into the web service had not been restarted following the upgrade. This has now been done.

• New observation format documentation

  2004 Mar. 11: 14:20. A slightly modified version of the documentation on the new observation format has been put on the anon-ftp site. The changes from the previous version are the definition of the format for occultation records and the cometary-magnitude aperture flag.
  • The .PS and .PDF versions of this document appear to be truncated at the end of page 12. It appears that earlier versions of this document also had this problem. We are investigating.
  • 14:27. Complete 15-page .PS and .PDF versions are now on-line.

• NEOCP start date

  2004 Mar. 9: 10:27. A user reported a problem with selecting an offset of -5 hours in the NEOCP, specifically that the entered offset was ignored. This problem was tracked down to a problem with calling a recently-extended library routine. New code has been put on-line.

• Webserver downtime

  2004 Mar. 8: 09:36. The webserver machine is about to be upgraded to the new OS version. The downtime should be about two hours.
  • The interface card has been replaced, but the CD-ROM is still not accessible. OS upgrade has been postponed. Service has been called. Upgrade will occur after CD-ROM drive has been replaced.
  • 14:30. Service is here and the CD-ROM is being replaced.
  • 15:00. CD-ROM has been replaced. Since our VMS consultant has to leave early today, the upgrade of SCULLY has been put off until tomorrow morning.
  • Mar. 9: 10:28. The upgrade of SCULLY is beginning. Downtime should be around two hours.

• March 6 mini-batch MPC files

  2004 Mar. 6: 12:00. A user reported that the files associated with the Mar. 6 mini-batch MPC were not updated correctly. A problem was detected in one command procedure used by the updating routines, concerning case-sensitivity of filenames. The affected routine has been made case-insensitive and the files are being regenerated.
  • 18:00. The correct files should now be in place.

• March 5 DOU MPEC

  2004 Mar. 5: 08:59. One of the upgraded machines seems to have a problem accessing the CF's webserver machine. By chance, the DOU MPEC was prepared on an execution queue running on this machine. Various bits involving copying of data to the CF's webserver failed. The majority of these failed bits have been rerun. Some problems may remain.

• Computer Upgrades V and the web services

  2004 Mar. 4: 15:56. One of the two remaining machines has been upgraded. The other, the webserver, has a faulty CD-ROM drive--service has been called, they won't be able to get here until tomorrow. The cluster is back and working. We are attending to a number of upgrade-related issues before reactivating the AUTOACK (and related) queues.
  • 17:01. The AUTOACK (and related) queues have been restarted.
  • 17:01. SCULLY will be upgraded sometime in the next few days.
  • 18:25. The web services have been reactivated. Please report any problems to the usual place.
  • Mar. 5: 10:00. Service arrived, will not replace SCSI card controlling CD-ROM as it is not one of theirs. We are obtaining a replacement interface card and will install it on Monday. Assuming that this fixes the problem of the non-functioning CD-ROM drive, the upgrade of the OS on SCULLY will then follow immediately.
  • Mar. 6: 10:00. Replacement interface card has been delivered overnight FedEx. Upgrade will occur on the morning on Mar. 8.

• Computer Upgrades IV

  2004 Mar. 4: 08:56. The upgrades of the final two machines, including the web server SCULLY, in the original cluster is beginning. The web services will probably be unavailable until both machines are upgraded, probably around noon.

• MPES oddities

  2004 Mar. 4: 08:56. Some users have reported oddities with obtaining ephemerides of numbered minor planets in the MPES. These problems are related to the ongoing computer upheaval. When the final machines have been upgraded, we will troubleshoot this problem.

• AUTOACK and MPES II

  2004 Mar. 3: 17:47. We've finished upgrading three of our cluster machines. We have restored internal and external (web) services. Two more machines remain to be upgraded: this will occur tomorrow morning. One of the remaining machines is the web server, SCULLY. Some further disruption is expected to occur at this time.
  • As noted below, tonight's DOU MPEC has been submitted to an execution queue.

• AUTOACK and MPES I

  2004 Mar. 3: 11:35. We've just discovered that in our attempt to standardize the list of authorized users on all machines in the cluster that we inadvertently erased five special accounts on SCULLY that are needed for, amongst other things, AUTOACK to function! We are currently adding these special accounts back into the list of authorized users.
  • The erroneously-deleted users have been reinstated.
  • However, it appears that it will be very difficult to maintain AUTOACK in a running state while the upgrades are proceeding. We are thus compressing the schedule for the upgrades in order to complete them as soon as possible.
  • The AUTOACK procedure will accept new incoming observation batches but will not ACK them automatically until the upgrades are complete. We will invoke the ACK mechanism manually at odd intervals until this time.
  • The web services are likely to be intermittent until the upgrades are complete, presumably some time tomorrow. Please be patient.

• March 3 DOU MPEC

  2004 Mar. 2: 22:45. Due to e-mail problems mentioned below, we are forced to cancel the issuance of tonight's DOU MPEC. We hope to return to normal service tomorrow night.
  • Mar. 3: 12:05. We may not be able to resume DOU MPEC production until the upgrades are completed.
  • Mar. 3: 17:50. Following the completion of today's upgrades a decision was made to submit tonight's DOU MPEC preparation job into the execution queue, in the expectation that it will complete successfully.

• Computer Upgrades III

  2004 Mar. 2: 09:16. The upgrades of the machines in the original cluster to the new version of the OS begin today. The webserver machine, SCULLY, is currently scheduled for the upgrade in the afternoon of March 3. The web cgi scripts that run on SCULLY will be unavailable during the upgrade of this machine. The scripts may also have problems if they need data off a disk that is being served to the cluster by another machine that is being upgraded.
  • Mar. 2: 11:56. The upgrade of SENECA is about to begin.
  • Mar. 2: 14:00. The upgrade of SENECA has finished. Some e-mail problems with the new setup have emerged. We are working on fixing them.
  • Mar. 2: 22:45. The e-mail problems persist. It appears we need to install a patch to fix them.
  • Mar. 3: 09:30. Reexamining the e-mail problem. It appears to affect only the user whose home directory resides on a disk attached to SENECA. Other users e-mailing from SENECA are apparently unaffected.
  • Mar. 3: 10:13. The e-mail problem has been tracked down and fixed. E-mail destined for the affected user is now being delivered.

Index to the CBAT/MPC/ICQ pages. Credits

MPC homepage