NERSCPowering Scientific Discovery for 50 Years

2005 User Survey Results

Services and Communications

 

  • Legend
  • Satisfaction with NERSC Services
  • How Important are NERSC Services to You?
  • How useful are these methods for keeping you informed?
  • Are you well informed of changes?
  • Comments about Services and Communications

 

 

Legend:

SatisfactionAverage Score
Very Satisfied 6.50 - 7.00
Mostly Satisfied 5.50 - 6.49
ImportanceAverage Score
Very Important 2.50 - 3.00
Somewhat Important 1.50 - 2.49
Significance of Change
significant increase
not significant
UsefulnessAverage Score
Very Useful 2.50 - 3.00
Somewhat Useful 1.50 - 2.49

 

Satisfaction with NERSC Services

7=Very satisfied, 6=Mostly satisfied, 5=Somewhat satisfied, 4=Neutral, 3=Somewhat dissatisfied, 2=Mostly dissatisfied, 1=Very dissatisfied

ItemNum who rated this item as:Total ResponsesAverage ScoreStd. Dev.Change from 2004
1234567
Account support services     1 1 4 25 119 150 6.73 0.61 0.06
Computer and Network Operations     1 2 1 24 64 92 6.61 0.73 0.10
Response to special requests (e.g. disk quota increases, etc.)       6 6 20 46 78 6.36 0.93 0.28
NERSC CVS server       2 3 10 13 28 6.21 0.92 0.88
Allocations process 1   5 5 7 49 57 124 6.16 1.11 0.23
E-mail lists     1 7 5 22 27 62 6.08 1.06 -0.04
Visualization services 1     7 3 9 16 36 5.83 1.42 0.43

 

How Important are NERSC Services to You?

3=Very important, 2=Somewhat important, 1=Not important

ItemNum who rated this item as:Total ResponsesAverage ScoreStd. Dev.
123
Allocations process 1 22 86 109 2.78 0.44
Account support services   39 92 131 2.70 0.46
Computer and Network Operations 2 27 59 88 2.65 0.53
Response to special requests (e.g. disk quota increases, etc.) 3 21 50 74 2.64 0.56
Visualization services 20 17 14 51 1.88 0.82
E-mail lists 21 27 13 61 1.87 0.74
NERSC CVS server 19 17 5 41 1.66 0.69

 

How useful are these methods for keeping you informed?

3=Very useful, 2=Somewhat useful, 1=Not useful

ItemNum who rated this item as:Total ResponsesAverage ScoreStd. Dev.
123
SERVICES: E-mail lists 1 42 104 147 2.70 0.47
MOTD (Message of the Day) 13 55 55 123 2.34 0.66
SERVICES: Announcements web archive 13 55 48 116 2.30 0.66
Phone calls from NERSC 25 25 32 82 2.09 0.83

 

Are you well informed of changes?

Do you feel you are adequately informed about NERSC changes?

AnswerResponsesPercent
Yes 161 99.4%
No 1 0.6%

Are you aware of major changes at least one month in advance?

AnswerResponsesPercent
Yes 137 85.6%
No 23 14.4%

Are you aware of software changes at least seven days in advance?

AnswerResponsesPercent
Yes 152 95.6%
No 7 4.4%

Are you aware of planned outages 24 hours in advance?

AnswerResponsesPercent
Yes 155 96.9%
No 5 3.1%

 

Comments about Services and Communications:   10 responses

MOTD / Communication of down times

MOTD is rarely useful b/c it mixes so many systems and is so long that the relevant pieces usually scroll off my screen before I see them. e.g. if PDSF logins would only show PDSF MOTD, that would be useful. ...

The MOTD seems never up to date.

I have noticed that the machine status does not always reflect reality. Several times I have been informed that Seaborg or Jacquard is "Up and available" when they are in fact down.

NERSC response: We are revamping the MOTD so it will be more concise and accurate. In particular, it will never show machines as being "Up and available" when they are not.

Authentication issues

... Account support: NERSC has very poor account support for collaborative computing (e.g. production accounts that are independent of individual user accounts). Currently essentially nothing exists for this and it has been frustratingly slow to get any action on this front. I consider this a major shortcoming of NERSC's computing framework.

How's about raising the number of incorrect login attempts before a lockout to, say, 20? If you make password requirements so strict, it's easy to forget the exact password, and if only a few guesses at my own password locks me out, then it's a waste of all our time to have it unlocked.

NERSC response: NERSC must abide by DOE password guidelines. These specify that "Three failed attempts to provide a legitimate password for an access request will result in an access lockout".

Other

It would be nice to have some kind of the mechanism to restore the jobs that has been crashed due to unexpected outages

NERSC response: This actually happens in many cases and users never know it. Sometimes this is not possible, however.

Enabling a PI to email everyone in their repo would be a great help.

NERSC response: We will investigate this.

Enabling a PI to hold queued/kill running jobs submitted by members of their repo would also help.

NERSC response: This is probably not possible for technical reasons. Among other issues: users can belong to multiple repos. PIs and Account Managers can limit the time available to each individual in their repo via the NIM accounting web interface.

Increase the maximum of time for a user code, if possible.

NERSC response: The longest wall times are on Seaborg, where the maximum wall time is 48 hours for 512 or more processors, and 24 hours for fewer than 512 processors, and Jacquard, where the maximum wall time is 48 hours for 2 to 32 processors and 24 hours for up to 64 processors. If this doesn't meet your needs, we urge you to contact the NERSC Users' Group, NUG. NUG has a significant influence on queue policies.

Have been very happy with extent to which personnel made contact directly to help us use NERSC most efficiently.

Generally satisfied with NERSC services. Applaud move to clusters. Allocation process is needlessly burdensome and machines are overallocated leading to long wait times

Given the sometimes dodgy nature of computer systems, I cannot think of anything that NERSC staff could do to keep us better informed of impending issues.