When a timshet (TCS metadata) failure occurs, what do I do?

1. We receive an email notifying us that hpf has lost tcs metadata. We may also hear an alert from netsounds.sh (although I have never heard anything).

2. Try to perform a graceful tims shutown by going to the rasession scripts window and issuing: "tims het shutdown"

3. If list item 2 is not successful, go to the tims client window (in rasession this is the upper-right window). In this window we enter a cntrl-C to kill the timshet run that is presumed to be hung. You can see more detailed information about the hpf scrrens at the top of the docement covering HPF. and also in a very detailed page HERE.

4. We hit the "TCS client restart" button in the RA Ops web to restart the timshet communication. We observe outputin the tims client window (rasession) that verities that the restart has occurred. --OR-- You can issue "timshet" in the rasession client window.

Later I found another description of this fix in the large HPF wiki doc. This doc has more detail and also and explanation of what to do to fix images where the tcs metadata has already been lost.