TO: All sites enabled to receive any of the HETE Notices 19 Jun 01 It is now known what caused the "burst" of HETE messages to the GCN system. A new field was added to the packets coming from the s/c. This field contained the attitude of the s/c. There is a filter in the HETE_OPS computer that looks for changes in the messages from the s/c. This filter is needed because multiple Secondary Ground Stations can receive a message (and forward it to HETE_OPS). This new field changes slightly (in the last few decimal places), so it looked like a "different" (ie unique) message from the s/c. So they were all forwarded to the GCN (and hense distributed to the world). HETE_OPS has changed their filtering software to ignore this new s/c attitude field. This type of scenario of multiple Notice distributes can not happen again. --scott ///////////////////////////////////////////////////////////////////////////// TO: All sites enabled to receive any of the HETE Notices 17 Jun 01 On Saturday 16 Jun 2001 the HETE spacecraft generated and sent to GCN 54 messages between 14:41:30 and 14:48:59 UT (delta_t=6.5min) for trigger #1554. And 12 messages between 16:25:13 and 16:26:48 (detla_t=1.5min) for trigger #1555. At this time, it is not known what caused the HETE flight software to generate so many messages. However, the GCN system had difficulty processing all the outgoing GCN Notices resulting from those incoming messages. The cpu loadfactor quickly exceeded 200 (the typical value during an episode of Notice distribution is 4). Under this load, most of the e-mail/pager/cellphone Notices did not get distributed -- certainly not the ones after the first 3 or 4 incoming messages. It is unclear how many of the socket sites received their packets. I have made a couple of changes to the system to mitigate against further occurances, but these changes are far from complete in their protection. By its vary nature, the GCN system has to be able to handle incoming messages from several sites simultaneously and to distribute the outgoing Notices to 100's of destinations. Asynchronous and autonomous operations preclude most protection measures. The HETE Operations center has informed me that they are temporarily stopping the connection between HETE and GCN until they can determine the cause of the problem (and hopely impliment a fix if needed). Scott Barthelmy GCN Operations