Discussion of Manchester WLCG Workshop agenda

Thursday 11 May 2017, 15:00 → 16:40 Europe/Zurich

513/R-068 (CERN)

Julia Andreeva (CERN)

 

15:00 → 15:15 Security and Trust session

Speaker: David Crooks

15:15 → 15:30 Network session

Speakers: Shawn Mc Kee (University of Michigan (US)), Tony Cass (CERN)

Three topics for the Networking session (2 hours)

1) Ask experiments about future Tier-N plans and network use  (15 minutes each experiment).   Scenarios to discuss with associated network "changes" anticipated:

·       Full—lots of cpu and disk more than adequate for caching needs, possibly able to support WAN access as well as LAN

·       CPU rich—i.e. some disk, but probably only enough to act as a local cache

·       Disk rich—if we imagine disk is still seen as hard to manage then maybe some T2s will specialise in this and maybe have less CPU than the average or way more disk. This could also be a configuration for an individual site in a distributed T2.

·       Disk poor—lower disk/cpu ratio than the CPU rich site

·       Diskless—is this feasible for a standalone T2? For  a site that is part of a distributed T2

2) Ask CMS to present on their work with scheduling pilots for I/O (network) intensive work.   Have contacted those working on this topic (no one planning to attend yet)

3) Discuss CWP networking topics  --- Review and discuss CWP network related items:  

     o  https://docs.google.com/document/d/1bOv5kDzmcg0eLSy7PZjjzE5OYSxaLAfaAxAI-IgKaac/edit

     o https://docs.google.com/document/d/11C9iCn3FPbvKBJpN0Lwv7tpjFELKPj14_dlkwuTGJqg/edit

Need to discuss with Monitoring and Operations which items to include from Networking.   Network Analytics, perfSONAR monitoring and maintenance, Alerting, etc.
 

15:30 → 15:45 Medium term evolution - Compute

Speaker: Gavin McCance (CERN)

Session 1: 90 mins  / COFFEE /  Session 2: 60 mins

Invited talks from each experiment to talk about medium term plans for their computing model. Aiming to bring out are opportunistic/HPC, mid-SLA, reducing ops costs (lightweight sites) and the potential uses of containers.

- ATLAS: Simone: yes, will find a someone
- ALICE: Latchezar can talk about current things and Run 3 planning
- CMS: no response yet
- LHCb : Andrew

---

1. Maarten has agreed to do an intro to the compute side of lightweight sites examine the possible models that people are thinking about. We could expand this with something on VAC/VCycle experience from GridPP.

2. Quite some desire to expose things on containers. Brian Bockelmann has offered a remote talk (-6GMT) on Singularity. Brian has also suggested some site-contacts. Haven't yet asked, but the talk on running compute workloads inside containers / kubernetes from RAL (at GridPP) by Andrew Lahiff was very interesting and I think would spark some discussions.  -> Links to security for docker/singlularity - what do we allow to run?

3. I could do my special on mid-SLA and using disk-servers.

---

CWP input with Peter to be decided with other sessions.

 

15:45 → 16:05 Medium term evolution - Data

Speaker: Oliver Keeble (CERN)

Topics are under discussion in the data steering group.

We'll certainly have something on caching storage and cloud/object stores.

We will present some results from the steering group (priorities, recommendations)

Still trying to understand the connection to CWP which is moving rather slowly for data.

16:05 → 16:20 Monitoring

Speaker: Alberto Aimar (CERN)

Monitoring Session (60 min)

MONIT at CERN for WLCG - status and plans 

Presentation (with small demos inside)

• General overview
• Status for 2017 (overlap with the current monitoring)
• Progressive migration (2018)

Status of WLCG

• FTS, Xrootd, transfers
• Job Monitoring
• Site monitoring, SSB
• LHC OPN

Other MONIT usage 

• DC monitoring, collectd
• IT Services (openstack, DNS, twiki)
• Logs and Timber

Monitoring at WLCG Sites  (a summarizer?)

• Someone doing a quick summary?
• A survey what people are using and moving to?

 

 MONIT for  ATLAS (Mario) 

• Summary of the WGs *Job Monitoring
• Site monitoring
• ATLAS reporting
• ASAP
• (Other ATLAS monitoring?)

MONIT for CMS (Carl or Valentin) 

• Activities of the Monitoring Forum
• Job Moniitoring
• WM Archive
• Glideinwms
• (Other CMS monitoring?)

ALICE  Monitoring 

• Discuss with Maarten

LHCb Monitoring

• Discuss with Andrea

Monitoring at Experiments  (a summarizer?)

• Someone doing a quick summary?
• A survey what people are using and moving to?

Discussion 

16:20 → 16:40 Operations

Speaker: Jose Flix Molina (Centro de Investigaciones Energéti cas Medioambientales y Tecno)