Home » Groups » STAR Protected

Shift accounting, dues for Run 14 and findings

Updated on Mon, 2014-02-17 12:01. Originally created by jeromel on 2013-12-24 08:53.

Quick links


Dues

Past dues & credits

From the Shift accounting, dues for Run 13 and findings, we ended with the following shift additions:
  • Cracow (Institute of Nuclear Physics PAS) was short by one shift and pro-actively requested to cover for +1 in Run 14
  • Jammu asked to serve +1 shift in Run 14 to re-equilibrate their average shift coverage
  • Pusan appealed (shift accounting in Run 13 would reduce authors) and agreed to serve +1 shift in Run 14
  • Wayne State University had their authors further reduced - one author was restored and it was agreed that more shifts would be covered to account for at least 2 authors
  • Warsaw University of Technology requested +3 shifts to account for a possible run extension coverrage from the beginning (i.e. they need to plan and regroup travels now and may not be able to do it later if an extension occurs). Since the polling for extra dues is done to cover for this kind of "travel issue" cases, we accomodated.
This brought to a +4 shifts to account for in Run 14.

Other considerations
  • Yale requested an emeritus credit for one of its member - at the time of the accounting, we had not heard from the council chair (specifically polled) and have no records of an emeritus member and could not update the PhoneBook records according to this request and suggested to the council rep from Yale to follow-up with the council chair (as commuication and transparency need impovement).
  • We received a request from U-Texas to ensure one QA shift slot for our QA coordinator (Lanny Ray) so he can better assess the issues at hand when working from remote - the Shift sign-up committee agreed the request is valid and will pre-sign one slot (his was made before opening)
  • UW requested more shift dues on the day before the opening but did not specify the excess due requested - it was requested for them to provide precisions before 4 PM BNL time. We received feedback past 4 PM but had the time to implement the +1 request mnts before the opening. In future, such a request will be ignored as it adds undue stress and work at the opening.

Shift layout

STAR shifts begins January 21st with cosmic data taking shifts. The crew will be manned with two people only. There shall be no period coordinators and no training shift slots.
The run this year is 15 weeks and starts on February 3rd. Real data shifts in STAR will begin on February 11th. The first two weeks will not hold training shifts. QA shifts begin on February 18th.
The last shift ends on May 20th.

As usual, period coordinators are pre-assigned / pre-signed as selected by the Spokesperson office. The list of period coordinators should be as follows:

Date Period coordinator Note
Jan 21st - Jan 28th None (cosmic) STAR cosmic data taking shifts
Jan 28th - Feb 4th None (cosmic)
Feb 4th - Feb 11th None (too early) Official RHIC run begins
Feb 11th - Feb 18th Martin Codrington Offline QA shifts available and real data taking expected
Feb 18th - Feb 25th Martin Codrington
Feb 25th - Mar 4th Martin Codrington
Mar 4th - Mar 11th Olga Evdokimov
Mar 11th - Mar 18th Olga Evdokimov
Mar 18th - Mar 25th Olga Evdokimov
Mar 25th - Apr 1st Salvatore Fazio
Apr 1st - Apr 8th Salvatore Fazio
Apr 8th - Apr 15th Shusu Shi
Apr 15th - Apr 22nd Shusu Shi
Apr 22nd - Apr 29th Shusu Shi
Apr 29th - May 6th Hao Qiu
May 6th - May 13th Hao Qiu
May 13th - May 20th Hao Qiu
     


Initial Run 14 dues

The shift composition leads to 66 training slots and 216 non-training shift slots (but no period coordinators during the cosmic weeks). This leads to 243 effective shifts to cover and 277 possible shifters.

The dues are as showed in the attached document below.

Snapshot at Run 14 shift extensions

Since the sign-up needs to be closed and set in countdown mode between the first period and the re-opening (shift extension), the below snapshot shows the people on shift in the week of the shift re-opening. Note that in Run 14, the re-opeing was set before the official full shifts started.



New features in Run 14

In Shift accounting, dues for Run 13 and findings, the main findings and improvements were 
  • STAR members continue to come on shifts un-trained, creating problems for theit crew and STAR (late replacement need to be found, we had to escort shifters as they did not have the proper training, ...)
  • Institution AAA may have a sign-up time advantage over other institutions appearing randomly in the institution dropdown list
    • The two issues above were addressed in two phases
    • In Shift requirements, improvement for Run 14, the design of a Web service (provided by ITD) and interface to the training database was made so it could be integrated in the Run 14 sign-up - the idea is to flag signer not properly trained and inform the shift leader (and possibly the period coordinator) a week prior to the beginning of the shift. This query-service requires entering the BNL badge ID
    • The Run 14 sign-up will include additional steps to take advantages training but also allows to pre-select member and institution before the opening (count down) release the sign-up for prime time.
  • Clear message should be displayed- instead of "over-subscription limit reached" the suggestion was to replaced with a clear "institution has reached its quota" message.
    • Done
  • Shift dues could be showed a fractional number upon mouse-over. This would allow people to see how close to the round-off they are in case of shift extension
    • Not implemented

The signup will now shows as follows:
  • As before, the first step will be to enter a CAPTCHA - this was added in past years to prevent scripting based sign-up. An example of how it looks is showed below

     
  • The second step will now offer you the possibility for the interface to access your training record. The panel should look like

    This step is not mandatory but
    • if you decline, you will be driven to the countdown and not be able t- pre-select your name and institution. The two issues were bound to one another to create an incentive to use the new feature.
    • If you accept, the following menue would appear

    Do not forget to enyter your ID and click "Check ID" - this will verify that we do have a record on file with this ID and can proceed. If you miss-type, you will be informed and you could correct it before moving further along. If yu ave full shift qualifications, you should see something like this coming up

    End by clicking [Proceed to ShiftSignup] and you will be brought to the countdown. Needless to say that if this phase tells "Regular Shifter trainings: NO" you shouldct accrodingly and inquire of the proper training you should take to be serving on a STAR shift. STAR trainng information is maintained by on the STAR training GUV center Web page.


Run 14 opening statistics and top 15 and top 3


Place Time Person Institution
1 10:00:09
 Xin Dong Lawrence Berkeley National Laboratory
2 10:00:11 James Drachenberg Valparaiso University
3 10:00:12
 Jacob Shearer Creighton University
4 10:00:13
 Sevil Salur Lawrence Berkeley National Laboratory
5 10:00:13 Ernst Sichtermann Lawrence Berkeley National Laboratory
6 10:00:13
 Prashanth Shanmuganathan Kent State University
7 10:00:13
 Xu Sun Lawrence Berkeley National Laboratory
8 10:00:13
 David Tlusty Nuclear Physics Inst., Academy of Sciences
9 10:00:14 James Draper University of California - Davis
10 10:00:15 Hua Pei Central China Normal University
11 10:00:15 Xuan Li Temple University
12 10:00:15 Bhanu Sharma Panjab University
13 10:00:16 Xiaoping Zhang Tsinghua University
14 10:00:16 Zhiming Li Central China Normal University
15 10:00:16 John Cramer University of Washington



Findings

Past findings were addressed and considered in this year's sign-up. However, the introduction of new features may also bring new technical issues.

Logistic issues

Before opening
  • The process of updating our PhoneBook records remains lengthy and tedious - previous attempt to made management widely aware has lead to little improvements and in fact, 2014 had its share of problems
  • The process and the diverse steps needed were at least discussed in the STAR management meeting (that's an improvement) - from there on ...
    • Many council members DID NOT respond to the request to update the records of their constituents - late record updates created an offset in the number of STAR authors by 15 authors, creating a need to redo the shift dues and accounting all over again
    • One institution asked for an Emeritus status update for one of its member - we have never heard from the council chair nor could find relevant notes in regards of a new emeritus member and could not officially consider this request (no PhoneBook update was made)
  • The process of gathering the list of experts was somewhat OK:  It was made 3 weeks later than initially expected but the final list was less confusing
  • We had the usual share of "although instructions are sent, some council members did not read it and could not find the information and/or link"
    • Although this is a classic and no solutions were found in the past, perhaps making a public list of problems would help resolve this wareness/attention issue (?)
  • A user from the West coast requested extention of access time to the test interface ... this request came a day prior to shift opening while the test interface was available since the 24th of December. We did not consider this request. However, this be one more example of why the council and management should act far in advance in proviidng the needed information for an earlier opening.
After opening
  • On the re-opening, we had a few adjustement requests of shift swap (this was easy to do without risking losing a slot) and regrouping of institutional shifts.
  • Projections were run on on the 7th and the 17th of February - in both instances, the institutions projected to have a rduction of authors were USNA and U-Washington. The later requested a shift due of +1 in Run 14 but did not sign-up accordingly to ensure an author buy-back.
  • On the 17th of February, the extra shift requested at the extention was removed (while all institutions noted here were contacted, only the Institute of Nuclear Physics, PAS got back to us on the re-opening for an extra shift sign-up - WSU wish list was to add +3 to do an author buy back on the re-opening but the intended shift coverage did not happen) 
  • The council was requested to cast an opinion on the following: "institutional group member will no longer appear in the sign-up dropdown menu after a date > their leave date" - previous considerations were that they may still do a shift on behalf of the institution. This will need to be set for Run 15 shift signup.
  • ...


Technical issues


The issues below are for internal purposes (and improving further for next year).

Before the opening
  • We had one case of a user not able to login during the test phase - a comment at the bottom of this page indicating the access credentials of the sign-up interface.
  • The training records check requires a check of the name implying our PhoneBook is syncrhonized with BNL's records. This was not the case for "Manuel Calderon" (our PhoneBook) versus "Manuel Calderon de la Barca Sanchez" (BNL records). Our PhoneBook was corrected but this suggests a enhenced fuzzy match.
  • ...
At or after the opening
  • People who did not refresh the sign-up page may have bumped into an ordering issue - CAPTCHA would be asked after the count down ends (we should make sure to remind people this is the case)
  • Observation: the top 10 is now randomized in terms of insitution name - Yeah!!! (i.e. the pre-signing works as intended)
  • If a user signs for a shift already taken by another user, there is no longer a warning message displayed (this may confuse the person signing up) - we need to re-verify the worklfow
  • An issue observed in Run 12 re-appeared - I copy below the problem description "then"
    • We had a case of a slot taken but the name not being displayed (member ID 1140)+ an empty row in the institution table appeared after the signup opened
    • Impact assessement: None for the sign-up but large disruptions/distraction for administrator (entries fixed by hand)
    • Status: investigating - both issues fund to be related to members saved in the sign-up DB with institution ID=0 (however, member has an institution assigned)
    • Several slots appeared again as showing an institution but no member name - resolved
    • This is not the same case - we enabled the tracing of signing/unsigning at the shift re-opening (as there as been reports of "I was signed but cannot see my name") and the code follows a diferent logic and unfortunately, showed the slots signed by a member of agiven institition A (who then un-signed) in red as if the slot was permanently allocated to instutition A. This was only a visual effect (the database did record all shifts taken properly).
  • ...

Institutions feedback and requests

{this section will contains a record of institutions requests for the next year - so far, no requests made}



Groups: