WLCG Service Challenges : WLCG Service Challenges Over (view) & Out (look)
---
Jamie Shiers, CERN
Introduction : Introduction Some feedback from the Tier2 workshop
Responses to questionnaire submitted so far
Planning for future workshops / meetings / events
SC4: Throughput Phase Results
SC4: Experiment Production Outlook
An Exciting Job Opportunity!
The Machine : The Machine Some clear indications regarding LHC startup schedule and operation are now available
Press release issued last Friday
Comparing our actual status with ‘the plan’, we are arguably one year late!
We still have an awful lot of work to do
Not the time to relax!
Press Release - Extract : Press Release - Extract CERN confirms LHC start-up for 2007
Geneva, 23 June 2006. First collisions in the … LHC … in November 2007 said … Lyn Evans at the 137th meeting of the CERN Council ...
A two month run in 2007, with beams colliding at an energy of 0.9 TeV will allow the LHC accelerator and detector teams to run-in their equipment ready for a full 14 TeV energy run to start in Spring 2008
Service Challenge ’07?
The schedule announced today ensures the fastest route to a high-energy physics run with substantial quantities of data in 2008, while optimising the commissioning schedules for both the accelerator and the detectors that will study its particle collisions. It foresees closing the LHC’s 27 km ring in August 2007 for equipment commissioning. Two months of running, starting in November 2007, will allow the accelerator and detector teams to test their equipment with low-energy beams. After a winter shutdown in which commissioning will continue without beam, the high-energy run will begin. Data collection will continue until a pre-determined amount of data has been accumulated, allowing the experimental collaborations to announce their first results.
Slide6 : Sectors 7-8 and 8-1 will be fully commissioned up to 7 TeV in 2006-2007
The other sectors will be commissioned up to the field needed for de-Gaussing (1.2 TeV)
Initial operation will be at 900 GeV (CM) with a static machine (no ramp, no squeeze) to dedug machine and detectors and to give a significant sample of W and Z
Full commissioning up to7 TeV will be done in the winter 2008 shutdown
Breakdown of a normal year : Breakdown of a normal year
Conclusions : Conclusions All key objectives have been reached for the end of 2005 and installation is now proceeding smoothly.
Three quarters of the machine has been liberated for magnet installation and interconnect work is proceeding in 2 octants in parallel. Magnet installation is now steady at 25/wk . Installation will finish end March 2007. The machine will be closed in August 2007.
Every effort is being made to establish colliding beams before the end of 2007 at reduced energy. The full commissioning up to 7 TeV will be done during the 2008 winter shutdown ready for a Physics run at full energy in 2008.
WLCG Tier2 Workshop : WLCG Tier2 Workshop This was the 2nd SC4 Workshop with primary focus on “new Tier2s”
i.e. those not (fully) involved in SC activities so far
1-2 people obviously didn’t know this from responses
Complementary to Mumbai “Tier1” workshop
Attempted to get Tier2s heavily involved in:
Planning the workshop (content)
The event itself
Chairing sessions, giving presentations and tutorials, …
Less successful in this than hoped – room for improvement!
Workshop Feedback : Workshop Feedback >160 people registered and participated!
This is very large for a workshop – about same as Mumbai
Some comments related directly to this
Requests for more tutorials, particularly “hands-on”
Requests for more direct Tier2 involvement
Feedback sessions, planning concrete actions etc.
Your active help in preparing / defining future events will be much appreciated
Please not just the usual suspects…
Workshop Gripes : Workshop Gripes Why no visit to e.g. ATLAS?
Why no introduction to particle physics?
These things could clearly have been arranged
Why no suggestion in the meeting Wiki?
Tutorial Rating – 10=best : Tutorial Rating – 10=best
Workshop Rating : Workshop Rating
SC Tech Day, June 21 : The first dedicated Service Challenge meeting (apart from workshops) for more than a year
Still hard to get feedback from sites and input from experiments!
Only 3 sites provided feedback by deadline!
Lunchtime Tuesday - input received from FZK, RAL and NDGF
By 14:30, ASGC, RAL, FZK, TRIUMF, NDGF with promises from IN2P3 & INFN
‘alright on the night’ – additional valuable expt info over T1 / T2 workshops
Logical slot for next SC Tech meeting would be ~September – prior to CMS CSA06 and ATLAS fall production
Avoiding EGEE 2006, Long Island GDB, LHCC comprehensive review…
Week starting September 11th IT amphitheatre booked Wed 13th & Fri 15th SC Tech Day, June 21
Future Workshops : Future Workshops Suggest ‘regional’ workshops to analyse results of experiment activities in SC4 during Q3/Q4 this year – important to drill down to details / problems / solutions
A ‘global’ workshop early 2007 focussing on experiment plans for 2007
Another just prior to CHEP
Given the size of the WLCG collaboration, these events are likely to be BIG!
Few suitable meeting rooms at CERN – need to plan well in advance
Something like 2 per year? Co-locate with CHEP / other events where possible?
Quite a few comments suggesting HEPiX-like issues. Co-locate with HEPiX?
A one-size-fits-all event is probably not going to succeed…
Jan 23-25 2007, CERN : Jan 23-25 2007, CERN This workshop will cover: For each LHC experiment, detailed plans / requirements / timescales for 2007 activities.
Exactly what (technical detail) is required where (sites by name), by which date, coordination & follow-up, responsibles, contacts, etc etc. There will also be an initial session covering the status of the various software / middleware and outlook. Do we also cover operations / support?
From feedback received so far, looks like an explicit interactive planning session would be a good idea
Dates: 23 January 2007 09:00 to 25 January 2007 18:00 (whole week now booked)
Location: CERN, Room: Main auditorium
Do we need tutorials? If so, what topics? Who can help?
Other ideas? Expert panel Q&A? International advisory panel?
Sep 1-2 2007, Victoria, BC : Sep 1-2 2007, Victoria, BC Workshop focussing on service needs for initial data taking: commissioning, calibration and alignment, early physics. Target audience: all active sites plus experiments
We start with a detailed update on the schedule and operation of the accelerator for 2007/2008, followed by similar sessions from each experiment.
We wrap-up with a session on operations and support, leaving a slot for parallel sessions (e.g. 'regional' meetings, such as GridPP etc.) before the foreseen social event on Sunday evening.
Dates: 1-2 September 2007
Location: Victoria, BC, Canada, co-located with CHEP 2007
Workshops - Summary : Workshops - Summary Please help to steer these events in a direction that is useful for YOU!
If you haven’t submitted a response to the questionnaire, please do so soon! (June 2006)
http://cern.ch/hep-project-wlcgsc/q.htm
Service Challenges - Reminder : Service Challenges - Reminder Purpose
Understand what it takes to operate a real grid service – run for weeks/months at a time (not just limited to experiment Data Challenges)
Trigger and verify Tier-1 & large Tier-2 planning and deployment – - tested with realistic usage patterns
Get the essential grid services ramped up to target levels of reliability, availability, scalability, end-to-end performance
Four progressive steps from October 2004 thru September 2006
End 2004 - SC1 – data transfer to subset of Tier-1s
Spring 2005 – SC2 – include mass storage, all Tier-1s, some Tier-2s
2nd half 2005 – SC3 – Tier-1s, >20 Tier-2s – first set of baseline services
Jun-Sep 2006 – SC4 – pilot service
Autumn 2006 – LHC service in continuous operation – ready for data taking in 2007
SC4 – Executive Summary : We have shown that we can drive transfers at full nominal rates to:
Most sites simultaneously;
All sites in groups (modulo network constraints – PIC);
At the target nominal rate of 1.6GB/s expected in pp running
In addition, several sites exceeded the disk – tape transfer targets
There is no reason to believe that we cannot drive all sites at or above nominal rates for sustained periods.
But
There are still major operational issues to resolve – and most importantly – a full end-to-end demo under realistic conditions SC4 – Executive Summary
Nominal Tier0 – Tier1 Data Rates (pp) : Nominal Tier0 – Tier1 Data Rates (pp) Heat
A Brief History… : A Brief History… SC1 – December 2004: did not meet its goals of:
Stable running for ~2 weeks with 3 named Tier1 sites…
But more sites took part than foreseen…
SC2 – April 2005: met throughput goals, but still
No reliable file transfer service (or real services in general…)
Very limited functionality / complexity
SC3 “classic” – July 2005: added several components and raised bar
SRM interface to storage at all sites;
Reliable file transfer service using gLite FTS;
Disk – disk targets of 100MB/s per site; 60MB/s to tape
Numerous issues seen – investigated and debugged over many months
SC3 “Casablanca edition” – Jan / Feb re-run
Showed that we had resolved many of the issues seen in July 2005
Network bottleneck at CERN, but most sites at or above targets
Good step towards SC4(?)
SC4 Schedule : SC4 Schedule Disk - disk Tier0-Tier1 tests at the full nominal rate are scheduled for April. (from weekly con-call minutes…)
The proposed schedule is as follows:
April 3rd (Monday) - April 13th (Thursday before Easter) - sustain an average daily rate to each Tier1 at or above the full nominal rate. (This is the week of the GDB + HEPiX + LHC OPN meeting in Rome...)
Any loss of average rate >= 10% needs to be:
accounted for (e.g. explanation / resolution in the operations log)
compensated for by a corresponding increase in rate in the following days
We should continue to run at the same rates unattended over Easter weekend (14 - 16 April).
From Tuesday April 18th - Monday April 24th we should perform the tape tests at the rates in the table below.
From after the con-call on Monday April 24th until the end of the month experiment-driven transfers can be scheduled.
Dropped based on experience of first week of disk – disk tests
Excellent report produced by IN2P3, covering disk and tape transfers, together with analysis of issues.
Successful demonstration of both disk and tape targets.
SC4 T0-T1: Results : SC4 T0-T1: Results Target: sustained disk – disk transfers at 1.6GB/s out of CERN at full nominal rates for ~10 days
Result: just managed this rate on Good Sunday (1/10)
Easter Sunday: > 1.6GB/s including DESY : Easter Sunday: > 1.6GB/s including DESY GridView reports 1614.5MB/s as daily average
Concerns – April 25 MB : Concerns – April 25 MB Site maintenance and support coverage during throughput tests
After 5 attempts, have to assume that this will not change in immediate future – better design and build the system to handle this
(This applies also to CERN)
Unplanned schedule changes, e.g. FZK missed disk – tape tests
Some (successful) tests since …
Monitoring, showing the data rate to tape at remote sites and also of overall status of transfers
Debugging of rates to specific sites [which has been done…]
Future throughput tests using more realistic scenarios
SC4 – Remaining Challenges : SC4 – Remaining Challenges Full nominal rates to tape at all Tier1 sites – sustained!
Proven ability to ramp-up rapidly to nominal rates at LHC start-of-run
Proven ability to recover from backlogs
T1 unscheduled interruptions of 4 - 8 hours
T1 scheduled interruptions of 24 - 48 hours(!)
T0 unscheduled interruptions of 4 - 8 hours
Production scale & quality operations and monitoring
Monitoring and reporting is still a grey area
I particularly like TRIUMF’s and RAL’s pages with lots of useful info!
Disk – Tape Targets : Disk – Tape Targets Realisation during SC4 that we were simply “turning up all the knobs” in an attempt to meet site & global targets
Not necessarily under conditions representative of LHC data taking
Could continue in this way for future disk – tape tests but
Recommend moving to realistic conditions as soon as possible
At least some components of distributed storage system not necessarily optimised for this use case (focus was on local use cases…)
If we do need another round of upgrades, know that this can take 6+ months!
Proposal: benefit from ATLAS (and other?) Tier0+Tier1 export tests in June + Service Challenge Technical meeting (also June)
Work on operational issues can (must) continue in parallel
As must deployment / commissioning of new tape sub-systems at the sites
e.g. milestone to perform disk – tape transfers at > (>>) nominal rates?
This will provide some feedback by late June / early July
Input to further tests performed over the summer
Combined Tier0 + Tier1 Export Rates : Combined Tier0 + Tier1 Export Rates CMS target rates double by end of year
Mumbai rates – scheduled delayed by ~1 month (start July)
ALICE rates – 300MB/s aggregate (Heavy Ion running)
SC4 – Meeting with LHCC Referees : SC4 – Meeting with LHCC Referees Following presentation of SC4 status to LHCC referees, I was asked to write a report (originally confidential to Management Board) summarising issues & concerns
I did not want to do this!
This report started with some (uncontested) observations
Made some recommendations
Somewhat luke-warm reception to some of these at MB
… but I still believe that they make sense! (So I’ll show them anyway…)
Rated site-readiness according to a few simple metrics…
We are not ready yet!
Disclaimer : Disclaimer Please find a report reviewing Site Monitoring and Operation in SC4 attached to the following page:
https://twiki.cern.ch/twiki/bin/view/LCG/ForManagementBoard
(It is not attached to the MB agenda and/or Wiki as it should be considered confidential to MB members).
Two seconds later it was attached to the agenda, so no longer confidential…
In the table below tentative service levels are given, based on the experience in April 2006. It is proposed that each site checks these assessments and provides corrections as appropriate and that these are then reviewed on a site-by-site basis.
(By definition, T0-T1 transfers involve source&sink)
Observations : Observations Several sites took a long time to ramp up to the performance levels required, despite having taken part in a similar test during January. This appears to indicate that the data transfer service is not yet integrated in the normal site operation;
Monitoring of data rates to tape at the Tier1 sites is not provided at many of the sites, neither ‘real-time’ nor after-the-event reporting. This is considered to be a major hole in offering services at the required level for LHC data taking;
Sites regularly fail to detect problems with transfers terminating at that site – these are often picked up by manual monitoring of the transfers at the CERN end. This manual monitoring has been provided on an exceptional basis 16 x 7 during much of SC4 – this is not sustainable in the medium to long term;
Service interventions of some hours up to two days during the service challenges have occurred regularly and are expected to be a part of life, i.e. it must be assumed that these will occur during LHC data taking and thus sufficient capacity to recover rapidly from backlogs from corresponding scheduled downtimes needs to be demonstrated;
Reporting of operational problems – both on a daily and weekly basis – is weak and inconsistent. In order to run an effective distributed service these aspects must be improved considerably in the immediate future.
Recommendations : Recommendations All sites should provide a schedule for implementing monitoring of data rates to input disk buffer and to tape. This monitoring information should be published so that it can be viewed by the COD, the service support teams and the corresponding VO support teams. (See June internal review of LCG Services.)
Sites should provide a schedule for implementing monitoring of the basic services involved in acceptance of data from the Tier0. This includes the local hardware infrastructure as well as the data management and relevant grid services, and should provide alarms as necessary to initiate corrective action. (See June internal review of LCG Services.)
A procedure for announcing scheduled interventions has been approved by the Management Board (main points next)
All sites should maintain a daily operational log – visible to the partners listed above – and submit a weekly report covering all main operational issues to the weekly operations hand-over meeting. It is essential that these logs report issues in a complete and open way – including reporting of human errors – and are not ‘sanitised’. Representation at the weekly meeting on a regular basis is also required.
Recovery from scheduled downtimes of individual Tier1 sites for both short (~4 hour) and long (~48 hour) interventions at full nominal data rates needs to be demonstrated. Recovery from scheduled downtimes of the Tier0 – and thus affecting transfers to all Tier1s – up to a minimum of 8 hours must also be demonstrated. A plan for demonstrating this capability should be developed in the Service Coordination meeting before the end of May.
Continuous low-priority transfers between the Tier0 and Tier1s must take place to exercise the service permanently and to iron out the remaining service issues. These transfers need to be run as part of the service, with production-level monitoring, alarms and procedures, and not as a “special effort” by individuals.
Site Readiness - Metrics : Site Readiness - Metrics Ability to ramp-up to nominal data rates – see results of SC4 disk – disk transfers [2];
Stability of transfer services – see table 1 below;
Submission of weekly operations report (with appropriate reporting level);
Attendance at weekly operations meeting;
Implementation of site monitoring and daily operations log;
Handling of scheduled and unscheduled interventions with respect to procedure proposed to LCG Management Board.
Site Readiness : Site Readiness 1 – always meets targets
2 – usually meets targets
3 – sometimes meets targets
4 – rarely meets targets
Site Readiness : 1 – always meets targets
2 – usually meets targets
3 – sometimes meets targets
4 – rarely meets targets Site Readiness
SC4 Disk – Disk Average Daily Rates : [1] The agreed target for PIC is 60MB/s, pending the availability of their 10Gb/s link to CERN. SC4 Disk – Disk Average Daily Rates
Site Readiness - Summary : Site Readiness - Summary I believe that these subjective metrics paint a fairly realistic picture
The ATLAS and other Challenges will provide more data points
I know the support of multiple VOs, standard Tier1 responsibilities, plus others taken up by individual sites / projects represent significant effort
But at some stage we have to adapt the plan to reality
If a small site is late things can probably be accommodated
If a major site is late we have a major problem
By any metrics I can think of, GridPP and RAL are amongst the best!
e.g. Setting up and testing T1-T2 transfers, active participation in workshops and tutorials, meeting technical targets, addressing concerns about 24x7 operations, pushing out info from WLCG MB etc.
I don’t think that this has anything to do with chance – THANKS A LOT!
Site Readiness – Next Steps : Site Readiness – Next Steps Discussion at MB was to repeat review but with rotating reviewers
Clear candidate for next phase would be ATLAS T0-T1 transfers
As this involves all Tier1s except FNAL, suggestion is that FNAL nominate a co-reviewer
e.g. Ian Fisk + Harry Renshall
Metrics to be established in advance and agreed by MB and Tier1s
(This test also involves a strong Tier0 component which may have to be factored out)
Possible metrics next:
June Readiness Review : June Readiness Review Readiness for start date
Date at which required information was communicated
T0-T1 transfer rates as daily average 100% of target
List the daily rate, the total average, histogram the distribution
Separate disk and tape contributions
Ramp-up efficiency (# hours, # days)
MoU targets for pseudo accelerator operation
Service availability, time to intervene
Problems and their resolution (using standard channels)
# tickets, details
Site report / analysis
Sites own report of the ‘run’, similar to that produced by IN2P3
Experiment Plans for SC4 : Experiment Plans for SC4 All 4 LHC experiments will run major production exercises during WLCG pilot / SC4 Service Phase
These will test all aspects of the respective Computing Models plus stress Site Readiness to run (collectively) full production services
These plans have been assembled from the material presented at the Mumbai workshop, with follow-up by Harry Renshall with each experiment, together with input from Bernd Panzer (T0) and the Pre-production team, and summarised on the SC4 planning page.
We have also held a number of meetings with representatives from all experiments to confirm that we have all the necessary input (all activities: PPS, SC, Tier0, …) and to spot possible clashes in schedules and / or resource requirements. (See “LCG Resource Scheduling Meetings” under LCG Service Coordination Meetings).
fyi; the LCG Service Coordination Meetings (LCGSCM) focus on the CERN component of the service; we also held a WLCGSCM at CERN last December.
The conclusions of these meetings has been presented to the weekly operations meetings and the WLCG Management Board in written form (documents, presentations)
See for example these points on the MB agenda page for May 24 2006
The Service Challenge Technical meeting (21 June IT amphi) will list the exact requirements by VO and site with timetable, contact details etc.
Summary of Experiment Plans : Summary of Experiment Plans All experiments will carry out major validations of both their offline software and the service infrastructure during the next 6 months
There are significant concerns about the state-of-readiness (of everything…) – not to mention manpower at ~all sites + in experiments
I personally am considerably worried –- seemingly simply issues, such as setting up LFC/FTS services, publishing SRM end-points etc. have taken O(1 year) to be resolved (across all sites).
and [still] don’t even mention basic operational procedures
And all this despite heroic efforts across the board
But – oh dear – your planet has just been blown up by the Vogons
[ So long and thanks for all the fish]
DTEAM Activities : DTEAM Activities Background disk-disk transfers from the Tier0 to all Tier1s will start from June 1st.
These transfers will continue – but with low priority – until further notice (it is assumed until the end of SC4) to debug site monitoring, operational procedures and the ability to ramp-up to full nominal rates rapidly (a matter of hours, not days).
These transfers will use the disk end-points established for the April SC4 tests.
Once these transfers have satisfied the above requirements, a schedule for ramping to full nominal disk – tape rates will be established.
The current resources available at CERN for DTEAM only permit transfers up to 800MB/s and thus can be used to test ramp-up and stability, but not to drive all sites at their full nominal rates for pp running.
All sites (Tier0 + Tier1s) are expected to operate the required services (as already established for SC4 throughput transfers) in full production mode.
(Transfer) SERVICE COORDINATOR
ATLAS : ATLAS ATLAS will start a major exercise on June 19th. This exercise is described in more detail in https://uimon.cern.ch/twiki/bin/view/Atlas/DDMSc4, and is scheduled to run for 3 weeks.
However, preparation for this challenge has already started and will ramp-up in the coming weeks.
That is, the basic requisites must be met prior to that time, to allow for preparation and testing before the official starting date of the challenge.
The sites in question will be ramped up in phases – the exact schedule is still to be defined.
The target data rates that should be supported from CERN to each Tier1 supporting ATLAS are given in the table below.
40% of these data rates must be written to tape, the remainder to disk.
It is a requirement that the tapes in question are at least unloaded having been written.
Both disk and tape data maybe recycled after 24 hours.
Possible targets: 4 / 8 / all Tier1s meet (75-100%) of nominal rates for 7 days
ATLAS Rates by Site : ATLAS Rates by Site ~25MB/s to tape, remainder to disk
ATLAS Preparations : ATLAS Preparations
ATLAS ramp-up - request : ATLAS ramp-up - request Overall goals: raw data to the Atlas T1 sites at an aggregate of 320 MB/sec, ESD data at 250 MB/sec and AOD data at 200 MB/sec.
The distribution over sites is close to the agreed MoU shares.
The raw data should be written to tape and the tapes ejected at some point. The ESD and AOD data should be written to disk only.
Both the tapes and disk can be recycled after some hours (we suggest 24) as the objective is to simulate the permanent storage of these data.)
It is intended to ramp up these transfers starting now at about 25% of the total, increasing to 50% during the week of 5 to 11 June and 75% during the week of 12 to 18 June.
For each Atlas T1 site we would like to know SRM end points for the disk only data and for the disk backed up to tape (or that will become backed up to tape).
These should be for Atlas data only, at least for the period of the tests.
During the 3 weeks from 19 June the target is to have a period of at least 7 contiguous days of stable running at the full rates.
Sites can organise recycling of disk and tape as they wish but it would be good to have buffers of at least 3 days to allow for any unattended weekend operation.
ATLAS T2 Requirements : ATLAS T2 Requirements (ATLAS) expects that some Tier-2s will participate on a voluntary basis.
There are no particular requirements on the Tier-2s, besides having a SRM-based Storage Element.
An FTS channel to and from the associated Tier-1 should be set up on the Tier-1 FTS server and tested (under an ATLAS account).
The nominal rate to a Tier-2 is 20 MB/s. We ask that they keep the data for 24 hours so, this means that the SE should have a minimum capacity of 2 TB.
For support, we ask that there is someone knowledgeable of the SE installation that is available during office hours to help to debug problems with data transfer.
Don't need to install any part of DDM/DQ2 at the Tier-2. The control on "which data goes to which site" will be of the responsibility of the Tier-0 operation team so, the people at the Tier-2 sites will not have to use or deal with DQ2.
See https://twiki.cern.ch/twiki/bin/view/Atlas/ATLASServiceChallenges
CMS : CMS The goal of SC4 is to demonstrate the basic elements of the CMS computing model simultaneously at a scale that serves as realistic preparation for CSA06. By the end of June CMS would like to demonstrate the following coarse goals.
25k jobs submitted on the total of the LCG and the OSG sites
Roughly 50% analysis jobs and 50% production jobs
50% Tier-1 resources and 50% Tier-2 resources
90% success rate for grid submissions
Information is maintained in the ‘CMS Integration Wiki’
https://twiki.cern.ch/twiki/bin/view/CMS/SWIntegration
Site list & status is linked
https://twiki.cern.ch/twiki/bin/view/CMS/SWIntSC4SiteStatus
Milestones are linked
https://twiki.cern.ch/twiki/bin/view/CMS/SWIntSC4Mile
CMS SC4 Transfer Goals : CMS SC4 Transfer Goals Data Transferred from CERN to Tier-1 centers with PhEDEx through FTS
ASGC: 10MB/s to tape
CNAF: 25MB/s to tape
FNAL: 50MB/s to tape
GridKa: 20MB/s to tape
IN2P3: 25MB/s to tape
PIC:20MB/s to tape
RAL: 10MB/s to tape
Data Transferred from Tier-1 sites to all Tier-2 centers with PhEDEx through FTS or srmcp
10MB/s for worst connected Tier-2 sites
100MB/s for best connected Tier-2 sites
Data Transferred from Tier-2 sites to selected Tier-1 centers with PhEDEx through FTS or srmcp
10MB/s for all sites to archive data
CMS CSA06 : CMS CSA06 A 50-100 million event exercise to test the workflow and dataflow associated with the data handling and data access model of CMS
Receive from HLT (previously simulated) events with online tag
Prompt reconstruction at Tier-0, including determination and application of calibration constants
Streaming into physics datasets (5-7)
Local creation of AOD
Distribution of AOD to all participating Tier-1s
Distribution of some FEVT to participating Tier-1s
Calibration jobs on FEVT at some Tier-1s
Physics jobs on AOD at some Tier-1s
Skim jobs at some Tier-1s with data propagated to Tier-2s
Physics jobs on skimmed data at some Tier-2s
ALICE : ALICE In conjunction with on-going transfers driven by the other experiments, ALICE will begin to transfer data at 300MB/s out of CERN – corresponding to heavy-ion data taking conditions (1.25GB/s during data taking but spread over the four months shutdown, i.e. 1.25/4=300MB/s).
The Tier1 sites involved are CNAF (20%), CCIN2P3 (20%), GridKA (20%), SARA (10%), RAL (10%), US (one centre) (20%).
Time of the exercise - July 2006, duration of exercise - 3 weeks (including set-up and debugging), the transfer type is disk-tape.
Goal of exercise: test of service stability and integration with ALICE FTD (File Transfer Daemon).
Primary objective: 7 days of sustained transfer to all T1s.
As a follow-up of this exercise, ALICE will test a synchronous transfer of data from CERN (after first pass reconstruction at T0), coupled with a second pass reconstruction at T1. The data rates, necessary production and storage capacity to be specified later.
More details are given in the ALICE documents attached to the MB agenda of 30th May 2006.
Last updated 12 June to add scheduled dates of 24 July - 6 August for T0 to T1 data export tests.
Production Activities and Requirements by ALICE : Production Activities and Requirements by ALICE Patricia Méndez Lorenzo
(on behalf of the ALICE Collaboration)
Service Challenge Technical Meeting
CERN, 21st June 2006
Outline : Outline Main Purposes of this presentation
Present the goals to achieve for the next months
Explain the specific actions, timelines and requirements for the sites
Expose our open questions
Parts of the talk
PDC’06/SC4 outline
Principles of operation
Immediate plans
Open questions and conclusions
Principal Goals of the PDC’06 : Principal Goals of the PDC’06 Use and test of the deployed LCG/gLite baseline services
Stability of the Services are fundamental during the 3 phases of the DC’06
Test of the data transfer and storage services
2nd phase of the DC’06
The stability and support of the services have to be assured beyond these tests
Test of distributed reconstruction and calibration model
Integrate the use of Grid resources with other resources available to ALICE within one single VO interface for different Grids
Analysis of reconstructed data
Plans for DC’06 : Plans for DC’06 First phase (ongoing):
Production of p+p and Pb+Pb events
Conditions and samples agreed with PWGs
Data migrated from all Tiers to castor@CERN
Second phase:
Push-out to T1s (July)
Registration of RAW data: 1st reconstruction at CERN, 2nd one at T1 (August/September)
User analysis on the GRID (September-October)
ALICE Software: General Aspects : ALICE Software: General Aspects Own Task Queue and related services
Pull model service: a server holds a master queue of jobs and it is up to the CE that provides the CPU cycles
Use of the LCG RB for the job agent submission
AliEn software used as a general front-end
It offers a single interface for ALICE users into different Grids
Resources used transparently and independently of the middleware behind
Make full use of the provided resources
Let the Grids to do their work
Integration of the LCG services using high-level tools and APIs
Developers put a lot of abstraction effort into hiding the complexity and shielding the user from implementation changes
Job Submission Structure : Job Submission Structure Optimizer Computing
Agent RB CE WN Env OK? Die with grace Execs agent ALICE Job Catalogue VO-Box LCG User Job ALICE catalogues ALICE File Catalogue packman
Principles of Operation: VOBOX : Principles of Operation: VOBOX VOBOXES deployed at all T0-t1-T2 sites
Requested in addition to all standard LCG Services
Entry door to the LCG Environment
It runs specific ALICE software components
Independent treatment at all sites
Apart of the FTS, the rest of services have to be provided at all sites
Installation and maintenance entirely ALICE responsibility
Based on a regional principle
Set of ALICE experts perfectly located at all sites
Site related problems handled by site admins
LCG Service problems reported via GGUS
Principle of Operations: FTS and LFC : Principle of Operations: FTS and LFC FTS Service
Used for scheduled replication of data between computing centers
Lower level tool that underlies the data placement
Used as FTD plug-in
FTS has been implemented through the FTS Perl APIs
FTD running in the VOBOX as one of the ALICE services
LFC deployed at all sites
Good relation with the sites is mandatory
Prompt and active behavior of the sites is required for a good performance
Current Implementation of FTS : Current Implementation of FTS Task Transfer Queue Site A SiteB 1. Contains the list of transfers
2. The Origin and the Destination (A and B)
3. The SURLs Via FTS in LCG sites a) Via FTD, the sites pull the transfer from the central server b) FTD makes a “get” to obtain information of the request c) FTS makes the transfer. FTD is an homogeneous layer for all type of sites
FTS Plans : FTS Plans ALICE plans to perform the transfer tests in July 2006 as follows:
From 10th-16th July:
Increase the data rate to 800MB/s DAQ-T0-Tape
Increase the xrootd pool to 10 disk server
From 17th-23rd July:
1GB/s full DAQ-T0-tape test for one consecutive week
No other data access in parallel to the DAQ-T0 tests
From 24th-30th July:
Start the export to the T1 sites at 200MB/s from the WAN pool
Test the reconstruction part at 300MB/s from the Castor2-xrootd pool
From 31st July-6th August:
Run the full chain at 300MB/s DAQ-T0-tape+reconstruction+export from the same pool
FTS Tests: Strategy : FTS Tests: Strategy The main goal is to test the stability and integration of the FTS service within FTD
T0-T1: 7 days required of sustained transfer rates to all T1s
T1-T2 and T1-T1: 2 days required of sustained transfers to T2.
Interaction done through the FTS-FTD wrapper developed in ALICE
Types of transfers
T0-T1: Migration of raw data produced at T0 and 1st reconstruction also at T0
T1-T2 and T2-T1: Transfers of reconstructed data (T1-T2) and Monte Carlo production and AOD (T2-T1) for custodial storage
T1-T1: Replication of reconstructed and AOD data
FTD-FTS Wrapper : FTD-FTS Wrapper The wrapper has basically two main options
Submission
The user will have to specify the SURL in the origin and in the destination... And that`s all
It will extract the SEs at the origin and the destination
From the IS, the names of the sites will be extracted and also the FTS Endpoint
The proxies status will be also checked
The transfer will be performed
Retrieve
The tool retrieves the status of all transfers associated to the previous submission
Additional Features
The tool creates automatically a subdirectory where the Ids and the status of the transfers are stored
It allows the specification of a file containing several transfers
FTS: Transfer Details : FTS: Transfer Details T0-T1: disk-tape transfers at an aggregate rate of 300MB/s at CERN
Distributed according the MSS resources pledged by the sites in the LCG MoU:
CNAF: 20%
CCIN2P3: 20%
GridKA: 20%
SARA: 10%
RAL: 10%
US (one center): 20%
T1-T2: Following the T1-T2 relation matrix
At this level, the transfer rates are not the primordial test but the services performance.
FTS: Transfer Remarks : FTS: Transfer Remarks The transferred data will not be stored at the remote site
Considered durable files, it is up of the site to remove the files on disk when needed
The FTS transfers will not be synchronous with the data production
Transfers based on LFN NOT required
The automatic update of the LFC catalog is not required
ALICE will take care of the catalogues update
Basic Requirements:
ALICE FTS Endpoints at the T0 and T1
SRM-enabled storage with automatic data deletion if needed
FTS service at all sites
Support during the whole tests(and beyond)
Proposed ALICE T1-T2 connections : Proposed ALICE T1-T2 connections CCIN2P3
French T2, Sejong (Korea), Lyon T2, Madrid (Spain)
CERN
Capa Town (South Africa), Kolkata (India), T2 Federation (Romania), RMKI (Hungary), Athens (Greece), Slovakia, T2 Federation (Poland), Wuhan (China)
FZK
FZU (Czech Republic), RDIG (Russia), GSI and Muenster (Germany)
CNAF
Tier2 Federation
RAL
Birmingham
SARA/NIKHEF
PDSF
Houston
Open questions : Open questions We have some remarks to discuss before the FTS transfers begin:
How do we have to ask to the responsible to enable the ALICE channels and FTS Endpoints?
Is it a site-experiment negociation?
The SC team will discuss it before and provide it to ALICE?
How is going to work the support?
Can we already count with a team of persons for the services and also at the sites to contact in case of problems?
Direct with the experts? With the sites? Via GGUS?
Conclusions : Conclusions The ALICE PDC`06 ongoing
Good support of the sites and extremely good relation between developers and support
Fast reactions to find and solve the problems
The 2nd Phase will begin with the FTS transfers
This is one of the main goals of the PCD`06
The transfer rates will be stressed but it is fundamental to test the services and the support
Still some open questions regarding the T1-T2 associations and the support
The 3rd phase (analysis) at the end of the year
LHCb : LHCb Starting from July LHCb will distribute "raw" data from CERN and store data on tape at each Tier1.
CPU resources are required for the reconstruction and stripping of these data, as well as at Tier1s for MC event generation.
The exact resource requirements by site and time profile are provided in the updated LHCb spreadsheet that can be found on https://twiki.cern.ch/twiki/bin/view/LCG/SC4ExperimentPlans
“LHCb plans”.
(Detailed breakdown of resource requirements in Spreadsheet)
Slide73 : LHCb DC’06 Challenge (using the LCG production services):
Distribution of RAW data from CERN to Tier-1’s
Reconstruction/stripping at Tier-1’s including CERN
DST distribution to CERN & other Tier-1’s
Aim to start beginning July
LHCb Tier-1’s:
Slide74 : LHCb DC’06 Pre-production before DC’06 already ongoing
Event generation, detector simulation & digitisation
For 100M B-physics events+100M min bias:
~ 4.2 MSI2k.months will be required (over ~2-3 months)
~ 125 TB of storage: data originally accumulated on MSS at CERN for DC06 stage Production will go on throughout the year for a LHCb physics book due in 2007
Slide75 : LHCb DC’06 Distribution of RAW data from CERN
~ 125 TB of storage replicated over Tier-1 sites (each file on only one Tier-1)
CERN MSS SRM local MSS SRM endpoint
Reconstruction/stripping at Tier-1’s including CERN
~300 kSI2k.months needed to reconstruct & strip events
(need to delete input file to recons after processing at Tier-1’s)
Output:
rDST (from recons): ~70 TB (accumulate) at Tier-1 sites where produced, on MSS
DST (from stripping): ~2TB in total on disk SE
DST distribution to CERN & all other Tier-1’s
2 TB of DST will be replicated to a disk based SE at each Tier-1 & CERN
Activities should go in parallel
Slide76 : Reconstruction (CERN & Tier-1’s) Need about ~0.4 MSI2k months needed for this phase (June & July)
500 events per file
Assume a job 4k events per job (8 i/p files)
2 hours/job
3.6 GB/job input; 2 GB/job output
Output stored at Tier-1 where job runs
50k jobs
Comparison with computing model:
2-3 GB (single) input files - ~100k events
2-3 GB output data
Jobs duration ~36 hours
~30k recons jobs per month during data taking
Slide77 : Stripping (CERN & Tier-1’s) Need about <0.1 MSI2k months needed for this phase (run concurrently with recons)
Assume a job 40k events per job (10 i/p files)
~2 hours/job
20 GB input/job; 0.5 GB data output per job
output distributed to ALL Tier-1’s & CERN
5k jobs
Comparison with computing model:
Jobs duration ~36 hours
10 input rDST files + 10 input RAW files - ~1M events
50 GB data input per job; ~7GB data output per job
~3k stripping jobs per month during data taking
Summary : Summary By GridPP17, the “Service Challenge” programme per se will be over…
But there are still some major components of the service not (fully) tested…
The Tier2 involvement is still far from what is needed for full scale MC production and Analysis (no fault of Tier2s…)
All experiments have major production activities over the coming months and in 2007 to further validate their computing models and the WLCG service itself, which will progressively add components / sites
We are now in production!
Next stop: 2020
Job Opportunity : Job Opportunity Wanted: man for hazardous journey.
Low wages, intense cold, long months of darkness and constant risks.
Return uncertain.
Sounds positively attractive to me…
WLCG Service Challenges : WLCG Service Challenges Over & Out