OceanStore: An Architecture for Global-Scale Persistent Storage John Kubiatowicz, et al ASPLOS 2000
OceanStore Global scale information storage. Mobile access to information in a uniform and highly available way. Servers are untrusted. Caches data anywhere, anytime. Monitors usage patterns users, each with 10,000 files
[Rhea et al. 2003] OceanStore
Main Goals Untrusted infrastructure Nomadic data
Example applications Groupware and PIM Digital libraries Scientific data repository Personal information management tools: calendars, s, contact lists
Example applications Groupware and PIM Digital libraries Scientific data repository Challenges: scaling, consistency, migration, network failures
Storage Organization OceanStore data object ~= file Ordered sequence of read-only versions Every version of every object kept forever Can be used as backup An object contains metadata, data, and references to previous versions
Storage Organization A stream of objects identified by AGUID Active globally-unique identifier Cryptographically-secure hash of an application-specific name and the owner’s public key Prevents namespace collisions
Storage Organization Each version of data object stored in a B- tree like data structure Each block has a BGUID Cryptographically-secure hash of the block content Each version has a VGUID Two versions may share blocks
[Rhea et al. 2003] Storage Organization
Access Control Restricting readers: Symmetric encryption key distributed to allowed readers. Restricting writers: ACL. Signed writes. ACL for object chosen with signed certificate.
Location and Routing Attenuated Bloom FiltersBloom Filters Find 11010
Location and Routing Plaxton-like trees
Updating data All data is encrypted. A set of predicates is evaluated in order. The actions of the earliest true predicate are applied. Update is logged if it commits or aborts. Predicates: compare-version, compare-block, compare-size, search Actions replace-block, insert-block, delete-block, append
Application-Specific Consistency An update is the operation of adding a new version to the head of a version stream Updates are applied atomically Represented as an array of potential actions Each guarded by a predicate
Application-Specific Consistency Example actions Replacing some bytes Appending new data to an object Truncating an object Example predicates Check for the latest version number Compare bytes
Application-Specific Consistency To implement ACID semantic Check for readers If none, update Append to a mailbox No checking No explicit locks or leases
Application-Specific Consistency Predicate for reads Examples Can’t read something older than 30 seconds Only can read data from a specific time frame
Replication and Consistency A data object is a sequence of read-only versions, consisting of read-only blocks, named by BGUIDs No issues for replication The mapping from AGUID to the latest VGUID may change Use primary-copy replication
Serializing updates A small primary tier of replicas run a Byzantine agreement protocol. A secondary tier of replicas optimistically propagate the update using an epidemic protocol. optimistically Ordering from primary tier is multicasted to secondary replicas.
The Full Update Path
Deep Archival Storage Data is fragmented. Each fragment is an object. Erasure coding is used to increase reliability. Erasure coding
Introspection computation observation optimization Uses: Cluster recognition Replica management Other uses
Software Architecture Java atop the Staged Event Driven Architecture (SEDA) Each subsystem is implemented as a stage With each own state and thread pool Stages communicate through events 50,000 semicolons by five graduate students and many undergrad interns
Software Architecture
Language Choice Java: speed of development Strongly typed Garbage collected Reduced debugging time Support for events Easy to port multithreaded code in Java Ported to Windows 2000 in one week
Language Choice Problems with Java: Unpredictability introduced by garbage collection Every thread in the system is halted while the garbage collector runs Any on-going process stalls for ~100 milliseconds May add several seconds to requests travel cross machines
Experimental Setup Two test beds Local cluster of 42 machines at Berkeley Each with GHz Pentium III 1.5GB PC133 SDRAM 2 36GB hard drives, RAID 0 Gigabit Ethernet adaptor Linux SMP
Experimental Setup PlanetLab, ~100 nodes across ~40 sites 1.2 GHz Pentium III, 1GB RAM ~1000 virtual nodes
Storage Overhead For 32 choose 16 erasure encoding 2.7x for data > 8KB For 64 choose 16 erasure encoding 4.8x for data > 8KB
The Latency Benchmark A single client submits updates of various sizes to a four-node inner ring Metric: Time from before the request is signed to the signature over the result is checked Update 40 MB of data over 1000 updates, with 100ms between updates
The Latency Benchmark Update Latency (ms) Key Size Update Size 5% Time Median Time 95% Time 512b 4kB MB b 4kB MB Latency Breakdown PhaseTime (ms) Check0.3 Serialize6.1 Apply1.5 Archive4.5 Sign77.8
The Throughput Microbenchmark A number of clients submit updates of various sizes to disjoint objects, to a four- node inner ring The clients Create their objects Synchronize themselves Update the object as many time as possible for 100 seconds
The Throughput Microbenchmark
Archive Retrieval Performance Populate the archive by submitting updates of various sizes to a four-node inner ring Delete all copies of the data in its reconstructed form A single client submits reads
Archive Retrieval Performance Throughput: 1.19 MB/s (Planetlab) 2.59 MB/s (local cluster) Latency ~30-70 milliseconds
The Stream Benchmark Ran 500 virtual nodes on PlanetLab Inner Ring in SF Bay Area Replicas clustered in 7 largest P-Lab sites Streams updates to all replicas One writer - content creator – repeatedly appends to data object Others read new versions as they arrive Measure network resource consumption
The Stream Benchmark
The Tag Benchmark Measures the latency of token passing OceanStore 2.2 times slower than TCP/IP
The Andrew Benchmark File system benchmark 4.6x than NFS in read-intensive phases 7.3x slower in write-intensive phases
[Koloniari and Pitoura] Bloom Filters Compact data structures for a probabilistic representation of a set Appropriate to answer membership queries
Bloom Filters (cont’d) Query for b: check the bits at positions H 1 (b), H 2 (b),..., H 4 (b). back
[Kang et al. 2003] Site ASite BSite C V 0 : (x 0 ) V 1 : (x 1 ) write x V 0 : (x 0 ) V 4 : (x 4 ) V 5 : (x 5 ) V 2 : (x 2 ) write x V 3 : (x 3 ) write x Pair-Wise Reconciliation
Site ASite BSite C V0V0 V1V1 V2V2 V3V3 V4V4 H0H0 H0H0 H1H1 H0H0 H2H2 H0H0 H3H3 H0H0 H1H1 H2H2 H4H4 H i = hash (V i ) V5V5 H3H3 H5H5 H0H0 H1H1 H2H2 H4H4 Hash History Reconciliation back
[Mitzenmacher] Erasure Codes Message Encoding Received Message Encoding Algorithm Decoding Algorithm Transmission n cn n n back