Architecture to Scale. DONN ROCHETTE at Big Data Spain 2012

Session presented at Big Data Spain 2012 Conference
16th Nov 2012
ETSI Telecomunicacion UPM Madrid
www.bigdataspain.org
More info: http://www.bigdataspain.org/es-2012/conference/architecture-to-scale/donn-rochette

Transcript

1. Big Data Spain 2012 The International Big Data Conference in Spain Madrid, 16th Nov 2012 ETSI Telecomunicacion UPM www.bigdataspain.org
2. Architecture for Scale A Case Study
3. Who am I? • CTO and Co-Founder of AppFirst • Application Virtualization o UNIX server applications o Solaris 2.6 applications on Solaris 10 • Real-time Operating Systems o Hubble Space Telescope o Under Wing Armaments o Medical Instruments • Launch Processing System o NASA Kennedy Space Center o Hardware and Software Design of Ground-based Launch Control Systems
4. • NYC based software start-up • Application o Aggregate & summarize data from 10ks of remote servers o Provide information for web apps and APIs • A Few Metrics o 45k to 50k summaries per minute o GBs per remote server per day o TBs of new data daily o Query & retrieve information in < 100 MS o Data store for up to 1 year AppFirst Collects, Aggregates and Correlates Information from Production Applications
5. Simplified Architecture
6. Design for Scale •Micro scale oApplication Components •Macro scale oThe Entire Service
7. Micro Scale: Data Processing Requirements: •Process a constant stream of data o3 snapshots per minute, per remote server •Create summaries in real-time oup to 1 minute behind wall clock time •Provide query results in < 100 MS
8. Micro Scale: Efficiency We found that: •Summaries of the data were needed in order to keep queries < 100 MS oServer oProcess oProcess sets oTopology •Time series needed for each summary type oMinute oHour oDay We tried: •Flat files •Network file systems •Distributed file systems •Relational databases •NoSQL key-value store •Memory based SQL databases •Distributed shared memory
9. Tape is Dead Disk is Tape Flash is Disk RAM Locality is King Jim Gray Microsoft December 2006 Micro Scale: We learned the hard way
10. Micro Scale: Solution Aggregation: •HPC pipeline processing model •RAM based data model •Queues as message bus •Stateless processing •Adaptive control •Queries are fully abstracted Horizontal scale may require that you revisit your design
11. Micro Scale We all know we need to scale horizontally Stateless •Any data processing with any time constraint •Processes can be run on any server •Processes can be migrated •Multiple processes can be added as load varies •All data stored in distributed shared memory •Message passing between components •Send keys and not data Cluster •Use components that cluster •Don’t do backups, use replication •Redis, memcached, RabbitMQ, Hbase can be clustered •Postgresql & MySQl don’t really cluster
12. Macro Scale: Application Capacity Load: •Most significant load impact from remote servers •User interaction, APIs, and queries do not load the system as much as remote servers •Support 100, 1,000, 10,000, 100,000 remote servers Will a design that supports 10,000 remote servers scale to support 100,000 remote servers?
13. Infinite Scale •Paralyzes the design team •Fosters bad behavior •Unrealistic expectations •Developers forced to take unrealistic action •But... you don’t want to say no to the business •The whole purpose is to add users •When the business brings a customer with 10,000 servers you want to say; bring it on
14. Macro Scale: Capacity We started with a snapshot: •Supported 1000 remote servers •Micro scale results made it possible to scale out •fairly flexible application component design •Scale out to 10,000 remote servers •This is a financial calculation •Scaled out in linear fashion •Data processing •Storage •Started in linear fashion then determined actual requirements
15. Macro Scale Solution: The Pod Pod Architecture: •Segmented infrastructure along the lines of load sources •Create infrastructure to support specific load •Instantiate additional infrastructure with additional load •When a pod gets to 85-90% capacity spin out a new pod •Capacity of a pod is a financial calculation •Scale within a pod in 1000 server increments •Need to automate the deployment of a pod Pod 0 Pod 1
16. Adaptive Control •You can’t react fast enough •Scale out •Scale back •Migrate Metrics are king •Business metrics •Application metrics Time Series Data •Issues relate to a specific time •Complete state information for any given minute •Don’t know what info is needed before a problem occurs; all data every minute Don’t trust the data •Clocks are skewed •Encodings fail •Save all bad data & replay •Think defensive The Pod Rocks •Isolated •Distributed •Located where needed •Behind the firewall
17. Conclusions •Stateless Data oKey to horizontal scale •Disk is tape oRAM based design is critical, not optional •Cluster oUse components that cluster, not just master/salve •Design for infinite scale does not work •Pod approach is an answer for infinite scale
18. Thank You! Donn Rochette donn@appfirst.com www.appfirst.com