WLCG Operations and Tools TEG

Introduction

WLCG has created a number of Technical Evolution Groups with the following objective:

To reassess the implementation of the grid infrastructures that we use in the light of
the experience with LHC data, and technology evolution, but never forgetting the
important successes and lessons, and ensuring that any evolution does not disrupt
our successful operation.

For more info, please see these presentations:

• http://indico.cern.ch/conferenceDisplay.py?confId=106648
• http://indico.cern.ch/materialDisplay.py?contribId=2&sessionId=2&materialId=slides&confId=106649

Input for WG4 and WG5

Note: WG4 was formerly known as WG3.

The editors of WG4 and WG5 are requesting combined input for the WLCG Operations And Tools TEG.

Please supply your input following the template below. Input is welcome either via twiki (extending this page), or email to the editors ( wlcg-teg-ops-wg3wg5@cernNOSPAMPLEASE.ch )

The editors will classify the input according to the different areas of work, but please limit your scope to the following:

• Middleware operational requirements
  • Example topics: log files, error messages, manageability, robustness, system requirements, documentation
• Service and middleware configuration
  • Example topics: puppet, quattor, yaim...
• Service and middleware deployment
  • Example topics: Packaging, upgrade paths, release schedules, rollout policy
• Middleware distribution
  • Example topics: Distros - EMI, EPEL, UMD, OSG, repositories, OS support, source availability

Please submit your feedback by 11th Nov.

NB - the Operations And Tools TEG has other areas of work not covered here:

• Monitoring and metrics
• Support tools
• Underlying services
• Application software management
• WLCG operations and procedures

Input template

Remember that answers can refer to any or all of the areas of interest (middleware operations, configuration, deployment and distribution). There is no need to answer every question.

Input provided for (site, experiment...):

1. What is in use currently and works well?
2. Top problems
3. What takes the most effort?
4. What can be improved about operational procedures?
5. What can be dropped?
6. Strategic directions you would like to recommend
7. Other comments

Example input

Input provided for (site, experiment...):

1. What is in use currently and works well?
   • staged rollout in EGI
2. Top problems
   • yaim is too complex
   • repository proliferation
   • log file formats are inconsistent
3. What takes the most effort?
   • middleware debugging
4. What can be improved about operational procedures?
   • more responsive releases
5. What can be dropped?
   • yaim
6. Strategic directions you would like to recommend
   • middleware should be distributed with puppet modules
7. Other comments

-- OliverKeeble - 31-Oct-2011