1. MongoDB at the energy frontier
Valentin Kuznetsov, Cornell University
MongoNYC, May, 2012
Monday, May 21, 12 1

2. Outline
✤ CMS :: LHC :: CERN
✤ Data Aggregation System and MongoDB
✤ Experience
✤ Summary
Monday, May 21, 12 2

3. CMS :: LHC :: CERN
Large Hadron Collider located at CERN, Geneva, Switzerland
CMS is one of the 4 experiments to probe our knowledge of particle
interactions and search for a new physics
Monday, May 21, 12 3

4. CMS :: LHC :: CERN
Compact Muon Solenoid (CMS)
Monday, May 21, 12 4

5. CMS :: LHC :: CERN
Typical proton-proton collision in CMS detector
Monday, May 21, 12 5

6. CMS :: LHC :: CERN
✤ 40 countries, 172 institutions, more then 3000 scientists
✤ CMS experiment produces a few PB of real data each year and we
collect ~TB of meta-data
✤ CMS relies on GRID infrastructure for data processing and uses 100+
computing centers word-wide
✤ CMS software consists of 4M lines of C++(framework), 2M lines of
python (data management), plus Java, perl, etc.
✤ ORACLE, MySQL, SQLite, NoSQL
Monday, May 21, 12 6

7. Dilemma
GenDB
LumiDB
Data
Quality
Phedex How I can find
my data?
DBS
PSetDB
SiteDB
Overview
RunDB
Monday, May 21, 12 7

8. Motivations
✤ Users want to query different
data services without knowing Data Aggregation System
about their existence
✤ Users want to combine
RunSummary run DataQuality LumiDB
information from different data run, trigger, detector, ... trigger, ecal, hcal, ... lumi, luminosity, hltpath
run, run lumi
services lumi
Phedex DBS
block, MC id
GenDB
✤ Some users may have domain block, file, block.replica,
file.replica, se, node, ... site
run, file, block, site,
config, tier, dataset,
lumi, parameters, ....
generator, xsection,
process, decay, ...
knowledge, but they need to
site
query X services, using Y SiteDB Overview
pset
Parameter Set DB
site, admin, site.status, .. country, node, region, .. CMSSW parameters
interface and dealing with Z
data formats to get our data Service E
param1, param2, DC
Service ..
Service
param1, param2, .. B
Service
param1, param2, .. A
Service
param1, param2, ..
param1, param2, ..
Monday, May 21, 12 8

9. Implementation idea
✤ When we talk we may use different
languages (English, French, etc.) or
different conventions (pounds vs kg)
✤ In order to establish communication
we use translation, dictionary,
thesaurus
Monday, May 21, 12 9

10. Implementation idea
Monday, May 21, 12 10

11. Pros
✤ Separate data management from discovery service
✤ Data are safe and secure
✤ Pluggable architecture (new translations)
✤ Users never bother with interface, naming and schema conflicts, data-
formats, security policies
✤ Information is aggregated in a real-time over distributed services
✤ Data consistency checks for free
✤ DB and API changes are transparent for end-users
Monday, May 21, 12 11

12. Cons
✤ DAS does not own the data
✤ lots of writes/reads/translations
✤ Data-services are real bottleneck
✤ nothing is guaranteed, e.g. service can go down, no control of its
performance, requested data can be really large, etc.
✤ cache often and preemptive
MongoDB to rescue !!!
Monday, May 21, 12 12

13. Data Aggregation System
Invoke the same API(params)
Update cache periodically
DAS robot Fetch popular
queries/APIs
DAS DAS DAS DAS
mapping Map data-service cache merge Analytics
output to DAS
records
record query, API
call to Analytics
runsum mapping aggregator
lumidb
data-services
parser
DAS core
DAS web
plugins
phedex CPU core RESTful interface
server
DAS core UI
sitedb
dbs DAS Cache server
Monday, May 21, 12 13

14. Mapping DB
✤ Holds translation between user keywords and data-service APIs,
resolve naming conflicts, etc.
✤ city=Ithaca query translates into Google API call
{'das2api': [{'api_param': 'q', 'das_key': 'city.name', 'pattern': ''}],
'daskeys': [{'key': 'city', 'map': 'city.name', 'pattern': ''}],
'expire': 3600,
'format': 'JSON',
'params': {'output': 'json', 'q': 'required'},
'system': 'google_maps',
'url': 'http://maps.google.com/maps/geo',
'urn': 'google_geo_maps'}
Monday, May 21, 12 14

15. Analytics DB
✤ Keep tracks of user queries, data-service API calls
{'api': {'params': {'q': 'Ithaca', 'output': 'json'}, 'name': 'google_geo_maps'}, 'qhash':
'7272bdeac45174823d3a4ea240c124ec', 'system': 'google_maps', 'counter': 5}
✤ Used by DAS analytics daemons to pre-fetch “hot” queries
✤ ValueHotSpot look-up data by popular values
✤ KeyHotSpot look-up data by popular key
✤ QueryMaintainer to keep given query always in cache
Monday, May 21, 12 15

16. Caching DB
✤ Data coming out from data-service providers are translated into JSON
and stored into cache collection
✤ naming translation are performed at this level
✤ Data records from cache collection are processed on common key, e.g.
city.name, and merged into merge collection
cache collection merge collection
{'city': {'name': 'Ithaca',
'lat':42, 'lng':-76}} {'city': {'name': 'Ithaca',
'lat':42, 'lng':-76,
{'city': {'name': 'Ithaca',
'zip':14850}}
'zip':14850}}
Monday, May 21, 12 16

17. DAS workflow query
DAS DAS
core logging
✤ Query parser
parser
✤ Query DAS merge collection yes no
query
DAS merge
✤ Query DAS cache collection yes
query
DAS cache
no
✤ invoke call to data service DAS DAS query DAS
merge cache data-services Mapping
✤ write to analytics
Aggregator DAS
Analytics
✤ Aggregate results results
✤ Represent results on web UI or via Web UI
command line interface
Monday, May 21, 12 17

18. Example
Monday, May 21, 12 18

19. DAS QL & MongoDB QL
✤ DAS Query Language built on top of MongoDB QL; it represents
MongoDB QL in human readable form
✤ UI level:
block dataset=/a/b/c | grep block.size | count(block.size)
✤ DB level:
col.find(spec={‘dataset.name’:‘/a/b/c’}, fields=[block.size]).count()
✤ We enrich QL with additional filters (grep, sort, unique) and
implement set of coroutines for aggregator functions
Monday, May 21, 12 19

20. DAS & MongoDB
✤ DAS works with 15 distributed data-services
✤ their size vary, on average O(100GB)
✤ DAS uses 40 MongoDB collections
✤ caching, mapping, analytics, logging (normal, capped, gridfs cols)
✤ DAS inserts/deletes O(1M) records on a daily basis
✤ We operate on a single 64-bit Linux node with 8 CPUs, 24 GB of RAM
and 1TB of disk space, sharding were tested, but it is not enabled
Monday, May 21, 12 20

21. MongoDB benefits
✤ Fast I/O and schema-less database are ideal for cache implementation
✤ you’re not limited by key:value approach
✤ Flexible query language allows to build domain specific QL
✤ stay on par with SQL
✤ No administrative costs with DB
✤ easy to install and maintain
Monday, May 21, 12 21

22. MongoDB issues (ver 2.0.X)
✤ We were unable to directly store DAS queries into analytics collection,
due to the dot constrain, e.g. {‘a.b’:1}
✤ queries <=> storage format {‘key’:‘a.b’, ‘value’:1}
✤ Scons is not suitable in fully controlled build environment
✤ it removes $PATH/$LD_LIBRARY_PATH for compiler commands;
it forces to use -L/lib64. As a result we used wrappers.
✤ Uncompressed field names and limitation with pagination/
aggregation
✤ should be addressed in new MongoDB aggregation framework
Monday, May 21, 12 22

23. Tradeoffs
✤ Query collisions: DAS does not own the data and there is no
transactions, we rely on query status and update it accordingly
✤ Index choice: initially one per select key, later one per query hash
✤ Storage size: we compromise storage vs data flexibility vs naming
conventions
✤ Speed: we compromise simple data access vs conglomerate of
restrictions (naming, security policies, interfaces, etc.), but we tuning-
up our data-service APIs based on query patterns
Monday, May 21, 12 23

24. Results
✤ The service in production over one year
✤ Users authenticated via GRID certificates and DAS uses proxy server
to pass credentials to back-end services
✤ Single query request yields few thousand records and resolved within
few seconds
✤ Pluggable architecture allows to query your service(s)
✤ unit tests are done against public data-services, e.g. Google, IP
look-up, etc.
Monday, May 21, 12 24

25. NoSQL @ CERN
✤ MongoDB is used by other experiments at CERN
✤ logging, monitoring, data analytics
✤ MongoDB is not the only NoSQL solution used at CERN
✤ One size does not fit all
✤ CouchDB, Cassandra, HBase, etc.
✤ There is on-going discussion between experiments and CERN IT
about adoption of NoSQL
Monday, May 21, 12 25

26. Summary
✤ CMS experiment built Data Aggregation System as an intelligent
cache to query distributed data-services
✤ MongoDB is used as DAS back-end
✤ During first year of operation we did not experience any significant
problems
✤ I’d like to thank MongoDB team and its community for their constant
support
✤ Questions? Contact: vkuznet@gmail.com
✤ https://github.com/vkuznet/DAS/
Monday, May 21, 12 26

27. Back-up slides
Monday, May 21, 12 27

28. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
Monday, May 21, 12 28

29. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
Monday, May 21, 12 28

30. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
block dataset=/a/b/c
MongoDB spec
Mapping DB
holds
relationships
Monday, May 21, 12 28

31. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
block dataset=/a/b/c
MongoDB spec
Mapping DB Caching DB
holds holds
relationships service records
Monday, May 21, 12 28

32. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
block dataset=/a/b/c
MongoDB spec
Mapping DB Caching DB Merge DB
holds holds holds
relationships service records merged records
Monday, May 21, 12 28

33. From query to results
Data service
generator Aggreator
API Data service Merge
Query Aggreator
lookup generator results
Data service Aggreator
generator
block dataset=/a/b/c
MongoDB spec
Mapping DB Caching DB Merge DB
holds holds holds
relationships service records merged records
Monday, May 21, 12 28