E-Log for Operations Meeting
–
SWE (Clark, Hall, De Contreras) met with Operations (Erickson, Stanek, Schuh,
Hultquist, Van Hoover, Hoobler, Allen, Colocho) to discuss where we are and how
we will proceed with the MCC Operations E-Log implementation. We went
through the agenda that
1) Mike Stanek has not yet talked in detail with CBAF,
but will. He indicated that some Operators and EOIC’s
have looked at or used the Test E-Log facility to become familiar with how the E-Log
behaves.
2) William Colocho reported that he is working with Terri
Lahey on being able to host a browser on an NCD, but that there are memory size issues and perhaps X-term issues that need
resolution. Browsers are always difficult to use because the “features”
of available browsers are not always compatible.
We talked
about tcl/tk as a potentially less painful
alternative, but no decision was reached.
Theo has indicated that the current Jefferson Lab code is not structured
well for migration to SLAC, but it could be done.
3) Operations had requested a filter to remove extra
lines in the XML. Bob Hall has implemented a solution for this
problem. Similarly, he is implementing an individual authorized writer
access list for each of the separate logs.
4) We discussed the Access based Shift Summary in the
context of a separate tool producing a file that would be converted and loaded
into the log when complete and final. Mike indicated that their understanding
was that there would be a machine auto-entry made at the end of the shift that
would be a header without text, and that the real report would be entered
later. The late entry would go in with a time stamp tied to its entry
time and place plus an addition reference on the earlier auto-entry line from
shift end. Neither entry would be editable after entry.
Cheryl wanted
to know if a tool other than Access was ok.
Ops can select any standard tool they wish; any choice will require a
translator to get to the log XML protocol.
The current Access tool was convenient since it is already in use and
there is an Access to XML translator in the EPICS Collaboration library. Ops should make a selection on this tool
sometime soon, and we will look at the translator effort.
5) Entry editing came next. Peter & Zoe suggested
several ways to allow editing by manipulating how entries are made “visible” in
the E-Log. While the database entry and
time stamp reflect entry time, the log text could be manipulated to reflect
only the later edited entries, with the earlier entries being “hidden”. For example, they would like to be able to
copy the original entry into a buffer and make modifications, additions, and deletions to that entry and then enter
the amended text into the log with the later time stamp. Then the
original entry would be “hidden”, so that the later entry substituted for the original
entry. All such changes could only be accomplished
by the original entry writer.
From a
technical point of view, this kind of change requires substantial custom code
to manage the rules for manipulating, handling, and hiding log entries. The real penalty, however, is that this
custom code has to be added to each new Collaboration release down the road – a
substantial maintenance effort/expense.
Changes to date requested by Ops have been implemented in the base
system, however, changes to “edit” rules have been rejected by Jefferson Lab as
contrary to DOE requirements.
6) Cheryl asked when we were going to switch over to this
Log as a production tool. Roger indicated that it is not about a
transition date but when we are ready and have the right E-Log features and
procedures.
We set the next meeting for
next Tuesday at