Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Towards an Upgrade TDR: LHCb Computing workshop 18-22 May 2015 Introduction to Upgrade Computing Session Peter Clarke Many peoples ideas Vava, Conor,

Similar presentations


Presentation on theme: "1 Towards an Upgrade TDR: LHCb Computing workshop 18-22 May 2015 Introduction to Upgrade Computing Session Peter Clarke Many peoples ideas Vava, Conor,"— Presentation transcript:

1 1 Towards an Upgrade TDR: LHCb Computing workshop 18-22 May 2015 Introduction to Upgrade Computing Session Peter Clarke Many peoples ideas Vava, Conor, Mike Patrick, Marco, Concezio...

2 2 Process  Discussion document  Purpose to start the “slow start” process (TCP) of getting people to think and contribute  It is understood that it is hard to get attention at the beginning of Run-II  It is understood that there is always a danger of starting too early //svn.cern.ch/reps/lhcbdocs/Users/pclarke/notes/Discussion-TowardsComputungTDR  LHCb discussion list : LHCb-UpgradeComputing-Discussion@cern.chLHCb-UpgradeComputing-Discussion@cern.ch  This meeting  Organise that R&D happens in good time to produce reports in 2016 and 2017  Benefit from experience of Run-II  Write TDR

3 3 Schedule Proposed schedule:  Q4 2015: Roadmap (of how we are going to get to TDR) Q4 2015  Q4 2016: Published notes on 2015/2016 experience  Q2 2017 R&D results written up  Q4 2017 Computing TDR submitted  Q4 2018 Computing model finalised

4 4 Questions to answer before we can write TDR  Lots of thoughts on ~ real time processing and data reduction  Requires change of mindset  Requires lots of testing  Trigger rate  Bandwidth  Event size  Rate  See next talks  Storage  What do we keep as RAW data – how much can it be reduced ?  Replica policy & Development of Intelligent data management  todays talks  MDST  MC storage or regeneration  CPU  CPU is 80% for MC  FastMC  Optimisation  See tuesdays talks

5 5 Questions(2)  Distributed computing hierarchy  Flatten hierarchy  concentrate at fewer high efficiency centres  The answer is as much political as technical (regional funding  willingness to pay for resources in one place). So maybe it is out of our hands  De-couple tape services from Tier-1  Use few high cost-efficiency tape vaults somewhere (at least for archive)  Possibly commercial  Review of main applications  Designed for mu of 0.4 and adapted ever since.  Is there a “new branch” development (break with past)  Tuesday discucssion  Data preservation  Scale of task will increase  Obligations will increase  We may even have our first requests by then as it is greater than 5 years !

6 6 Questions(3): People model  Current people model for operations is  CERN centric  pressure on limited number of CERN based people  Clearly it worked for Run-I and presumably will for Run-II  But it is too reliant upon good will of small set or people  Must it be this way ? Is this what we want in Run-III  We may try to de-centralise  Pass specific responsibilities to external groups  Of course this has been tried before with limited success, but it is perhaps worth trying again.  We should survey the other experiments.

7 7 Immediate needs  All of these things (and more) need investigation  Work between now and ~ Q2 2017  Experience from Run-II  Some changes will be breaking changes  Need to plan when technical break occurs  Need to plan when psychological break occurs for collaboration members  Of course many of these areas are on-going work and its is important to recognise and appreciate this  Nevertheless I am personally convinced that it is essential to identify people to be “project managers” responsible for at least some of these topics  Writing down the work needed  timeline planning  report production  Of course I am not oblivious to the discussions earlier in the week on chronic lack of suitable effort  Maybe we need to divert people from Run-II anyway !  This may then have a detrimental effect on ability to deal with Run-II data but we may simply not have the choice if we are to be ready for 2020 !

8 8 Immediate needs  If there were team of people available in an institute then we would do this properly with project managers now. There are not  An upgrade detector has a dedicated set of engineers which are disjoint from running LHCb Run-II  In computing its the same people  Some things are in the category that they will go on anyway  E.g. Data Popularity (Intelligent Data Management)  Todays talk  Probably fine to expect progress and a report ready for TDR  Some things may be ok to only start after the TDR in LS2 (when effort becomes available)  Developing means to regenerate MC (and not store it)  But some things are just like a detector and need serious planned programme starting now  Event oriented processing  Full tests of highly reduced data sets.  FastMC ?  It is this last category which we do not have the >50% FTE project managers for  Will only happen as a secondary/best efforts task  May become too late for 2020

9 9 Immediate needs  Efforts need to be re-doubled to educate our collaboration members  Software is no more or less a professional task than building a detector  If the effort is not found – things will not be done – and Run-III will suffer  Group Leaders must be asked again (futile as it may be) to consider prioritising people in their groups to do software tasks.  Means need to be found to hire the missing project managers (or displacing existing people) by charging institutes or charging M&O otherwise  Continue all efforts to ensure software work is appreciated [this is complex]  Management needs to support this position so that it is not just Marco and I pleading.

10 10 Immediate needs  Efforts need to be re-doubled to educate our collaboration members  Software is no more or less a professional task than building a detector  If the effort is not found – things will not be done – and Run-III will suffer  Group Leaders must be asked again (futile as it may be) to consider prioritising people in their groups to do software tasks.  Means need to be found to hire the missing project managers (or displacing existing people) by charging institutes or charging M&O otherwise  Continue all efforts to ensure software work is appreciated [this is complex]  Management needs to support this position – so its not just Marco and I pleading LHCb Collaboration meeting Photo, 2019


Download ppt "1 Towards an Upgrade TDR: LHCb Computing workshop 18-22 May 2015 Introduction to Upgrade Computing Session Peter Clarke Many peoples ideas Vava, Conor,"

Similar presentations


Ads by Google