When a safing is announced, the LRPG must find out what the programatic impact is. To do that, we need to find out what visits have been and are going to be hit, as well as visits that potentially could be hit--as quickly as possible after the safing event starts.

1. Create a directory to store the analysis in.
   cd /data/implementation/www/html-data/spst/lrpg/safing/

   doy

   mkdir <yyddd>

   cd <yyddd>

2. Extract LRP and SPSS information on visits potentially impacted. This is done by running the tool
   create-safing-lists.bash

   which does the following:

   • Save a list of all the visits that might have been scheduled on the impacted week;
   • Separate mains, non-main visits (into pars/cals and SNAPs);
   • Extract the essentials from this list:
   • Match/join this list against the RAFV list.
   • trims the final 'lj' file to "just" 185 characters wide.

   This list is a join of the SPSS scheduling information and the LRP 'available' type information (without plan windows). Note that these are 1 column wider than the 'available' and RAFV lists in that the sunit_id is repeated within and additionally the c&c list name replaces the #ofPWlets column.

3. Edit the list to remove the visits which occurred before the day of the safing event.

   Note: Watch out for the Anneals and Decons. These do not pop up in the main science/cal list, and you will have to check the HST status page for their prop_ids: http://www.stsci.edu/org/ess/teams/fse/flight_stat?start=25 and grep through the pars/cals list (which is generated in parallel by the above tool) to see where the vistis are scheduled.

   For example, in April 2011, there are 3 such proposals, and we can extract them by:

   egrep '(12343|12387|12404)' schedrecently-parcals-exectimes-rafv.lj schedrecently-mains-exectimes-rafv.lj

   If they are impacted, the PC and SMSB team lead need to be aware since they will either need to be replanned or (more likely) rescheduled soon after the vehicle recovery time.

4. Start up a SPIKE image if one is not already available. Do a 'load props' and paste the proposal IDs from the unix command line that were output by the previous set of commands.

   There are some simplified ways of extracting the prop_id and sunit_ids. To get the prop_ids:

   getvorp -v -p -s <filename>

   To get the sunit_ids:

   getvorp -v -v <filename>

   Once the proposals are finished loading, display them and proceed with any required replan analysis.

Following the Impact Analysis Procedure creates many of the items that will be used for posting. The URLs are

• Behind the firewall:
  http://www.sogs.stsci.edu/spb/swat/safing/safe.html

• Outside the firewall:
  http://www.stsci.edu/spb/swat/safing/safe.html

1. Log into lrp on a unix box.

2. cd /data/implementation/www/html-data/spst/lrpg/safing/<yyddd>/

3. cp yyddd/safe.txt

   Place your copy of the safe mode notification mail into the correct directory with the correct name.

   SPECIAL NOTE: For all emails posted here, make certain that email addresses and telephone numbers have been removed. This is to ensure the privacy of senders in case this information is ever distributed outside of the firewall or the HST network.

4. cp yyddd/update#.txt

   Place your copy of the update mail into the correct directory with the correct name. # is the version of the update. Some safings have had 4 or 5 updates.

5. cp yyddd/lost.txt

   The file lost.txt can be created a number of different ways, but copying the data from the ".lj" files is probably most straightforward.

6. edit ../safe.html

   Edit the file to have the following format. The changes are put at the top of the file just below the header. As needed, create the title line, and the hooks to the three different types of files. E.g.:

   27 September 1997 (1997.270) NICMOS Suspend

   • Safing Notice
   • Update and recovery plans
   • Lost observations

7. Replace the dates, titles, etc. as needed. Also the date at the top of the file should be changed.

8. Finally, write out the changes, load the page via a browser and make sure it looks okay and all the links work. Do this from the site behind the firewall. Because changes sometimes can take hours to propagate to outside the firewall.