Sailesh Krishnamurthy
Palo Alto, California, United States
8K followers
500+ connections
Activity
-
It's pretty amazing what the Memorystore engineering team has accomplished for vector search performance and scale: a single Valkey or Redis Cluster…
It's pretty amazing what the Memorystore engineering team has accomplished for vector search performance and scale: a single Valkey or Redis Cluster…
Liked by Sailesh Krishnamurthy
-
Congratulations to my colleagues at Google workin on the RECAP+Gemini NL2SQL technique. The team's latest entry has taken back the top position in…
Congratulations to my colleagues at Google workin on the RECAP+Gemini NL2SQL technique. The team's latest entry has taken back the top position in…
Shared by Sailesh Krishnamurthy
-
Happy to share a set of announcements relating to Gemini and VertexAI. -- Model Updates: * New versions of Gemini 1.5 Pro and Flash with…
Happy to share a set of announcements relating to Gemini and VertexAI. -- Model Updates: * New versions of Gemini 1.5 Pro and Flash with…
Liked by Sailesh Krishnamurthy
Publications
-
Amazon Aurora: Design Considerations for High Throughput Cloud-Native Relational Databases
SIGMOD 2017
Amazon Aurora is a relational database service for OLTP workloads offered as part of Amazon Web Services (AWS). In this paper, we describe the architecture of Aurora and the design considerations leading to that architecture. We believe the central constraint in high throughput data processing has moved from compute and storage to the network. Aurora brings a novel architecture to the relational database to address this constraint, most notably by pushing redo processing to a multi-tenant…
Amazon Aurora is a relational database service for OLTP workloads offered as part of Amazon Web Services (AWS). In this paper, we describe the architecture of Aurora and the design considerations leading to that architecture. We believe the central constraint in high throughput data processing has moved from compute and storage to the network. Aurora brings a novel architecture to the relational database to address this constraint, most notably by pushing redo processing to a multi-tenant scale- out storage service, purpose-built for Aurora. We describe how doing so not only reduces network traffic, but also allows for fast crash recovery, failovers to replicas without loss of data, and fault-tolerant, self-healing storage. We then describe how Aurora achieves consensus on durable state across numerous storage nodes using an efficient asynchronous scheme, avoiding expensive and chatty recovery protocols. Finally, having operated Aurora as a production service for over 18 months, we share lessons we have learned from our customers on what modern cloud applications expect from their database tier.
-
Events on the Edge
SIGMOD
Demonstrated complex event detection in a heterogeneous receptor network built using RFID and motion sensors, deployed in a library scenario.
Patents
-
Systems and methods for managing queries
Issued US 8,903,802
A streaming database management system may provide support for creating a new continuous query operator for on-the-fly addition of a new query, e.g., a new SQL query and/or new continuous query operator, into a shared continuous query plan. Alternatively, the streaming database management system may associate the new query on-the-fly with an existing continuous query operator, such as a relational operator and expression evaluator, to reuse the existing continuous query operator. In some…
A streaming database management system may provide support for creating a new continuous query operator for on-the-fly addition of a new query, e.g., a new SQL query and/or new continuous query operator, into a shared continuous query plan. Alternatively, the streaming database management system may associate the new query on-the-fly with an existing continuous query operator, such as a relational operator and expression evaluator, to reuse the existing continuous query operator. In some embodiments, multiple operators are grouped for on-the-fly addition to the shared continuous query plan. Alternatively, the streaming database management system may identify a group of multiple operators in the shared continuous query plan for reuse by the new query.
Other inventors -
Addition and processing of continuous SQL queries in a streaming relational database management system
Issued US 8,745,070
Systems, methods, and media are disclosed herein that can be embodied in a traditional Relational Database Management System (RDBMS) in order to transform it into a Streaming Relational Database Management System (SRDBMS). An SRDBMS may provide functionality such as to manage and populate streams, tables, and archived stream histories and support the evaluation of continuous queries on streams and tables. Both continuous and snapshot queries support the full spectrum of the industry standard…
Systems, methods, and media are disclosed herein that can be embodied in a traditional Relational Database Management System (RDBMS) in order to transform it into a Streaming Relational Database Management System (SRDBMS). An SRDBMS may provide functionality such as to manage and populate streams, tables, and archived stream histories and support the evaluation of continuous queries on streams and tables. Both continuous and snapshot queries support the full spectrum of the industry standard, widely used, Structured Query Language. The present technology can support a high number of concurrent continuous queries using a scalable and efficient shared query evaluation scheme, support on-the-fly addition of continuous queries into a mechanism that implements the shared evaluation scheme, reuse RDBMS modules such as relational operators and expression evaluators, and visualize results of continuous queries in real time.
Other inventorsSee patent -
Order-Independent Stream Query Processing
Issued US 8484243
In a system and method for order-independent stream query processing, one or more input streams of data are received, and the one or more input streams are analyzed to determine data which is older than an already emitted progress indicator. The data which is older than the already emitted progress indicator is partitioned into one or more partitions, and each of the one or more partitions are independently processed using out-of-order processing techniques. A query is received, rewritten and…
In a system and method for order-independent stream query processing, one or more input streams of data are received, and the one or more input streams are analyzed to determine data which is older than an already emitted progress indicator. The data which is older than the already emitted progress indicator is partitioned into one or more partitions, and each of the one or more partitions are independently processed using out-of-order processing techniques. A query is received, rewritten and decomposed into one or more sub-queries that produce partial results for each of the one or more partitions, where each of the one or more sub-queries correspond to a partition. A view is also produced that consolidates the partial results for each partition. The partial results are consolidated at a consolidation time specified by the query to produce final results, and the final results are provided.
Other inventors -
System and method for adaptive database caching
Issued US 7,945,577
A local database cache enabling persistent, adaptive caching of either full or partial content of a remote database is provided. Content of tables comprising a local cache database is defined on per-table basis. A table is either: defined declaratively and populated in advance of query execution, or is determined dynamically and asynchronously populated on-demand during query execution. Based on a user input query originally issued against a remote DBMS and referential cache constraints between…
A local database cache enabling persistent, adaptive caching of either full or partial content of a remote database is provided. Content of tables comprising a local cache database is defined on per-table basis. A table is either: defined declaratively and populated in advance of query execution, or is determined dynamically and asynchronously populated on-demand during query execution. Based on a user input query originally issued against a remote DBMS and referential cache constraints between tables in a local database cache, a Janus query plan, comprising local, remote, and probe query portions is determined. A probe query portion of a Janus query plan is executed to determine whether up-to-date results can be delivered by the execution of a local query portion against a local database cache, or whether it is necessary to retrieve results from a remote database by executing a remote query portion of Janus query plan.
-
Method, system, and article of manufacture for transferring structured data between different data stores
Issued US 7,136,872
Provided is a method, system, and an article of manufacture for transferring structured data between different data stores. Data stores include a type identifier with every instance of a structured data type. A typeID-typeName map within each data store includes the mapping between the type identifiers and the type names of the structured data types. The typeID-typeName map is sent as metadata while transferring data from a source database to a target database. An application replaces the type…
Provided is a method, system, and an article of manufacture for transferring structured data between different data stores. Data stores include a type identifier with every instance of a structured data type. A typeID-typeName map within each data store includes the mapping between the type identifiers and the type names of the structured data types. The typeID-typeName map is sent as metadata while transferring data from a source database to a target database. An application replaces the type identifiers of the source database with the corresponding type identifiers of the target database
Other inventors -
Signature hash for checking versions of abstract data types
Issued US 6,973,572
A method, apparatus, and article of manufacture for providing to a signature hash for checking versions of abstract data types. An identifier is constructed for the abstract data type that is substantially unique to the abstract data type, wherein the identifier comprises a concatenation of various attributes for the abstract data type. The constructed identifier is hashed to generate a signature hash value for the abstract data type, which is then stored both in the database and a class…
A method, apparatus, and article of manufacture for providing to a signature hash for checking versions of abstract data types. An identifier is constructed for the abstract data type that is substantially unique to the abstract data type, wherein the identifier comprises a concatenation of various attributes for the abstract data type. The constructed identifier is hashed to generate a signature hash value for the abstract data type, which is then stored both in the database and a class definition for the abstract data type. When the class definition is instantiated as a library function, it accesses the abstract data type from the database, and compares the signature hash value from the database and the signature hash value from the class definition in order to verify that the class definition is not outdated. The class definition is outdated when the abstract data type has been altered without the signature hash value being re-generated and re-stored in the database and the class definition.
-
Method for providing for persistence of java classes where the persistence semantics may be orthogonal to the class definition
Issued US 6,505,211
Languages
-
Tamil
-
More activity by Sailesh
-
The Spanner team is on a mission to revolutionize the way businesses build and scale their applications. We're committed to providing a globally…
The Spanner team is on a mission to revolutionize the way businesses build and scale their applications. We're committed to providing a globally…
Liked by Sailesh Krishnamurthy
-
Faster to AI value with AlloyDB AI and machine learning are all the rage right now. In my last year as an analyst at Gartner, there was nothing…
Faster to AI value with AlloyDB AI and machine learning are all the rage right now. In my last year as an analyst at Gartner, there was nothing…
Liked by Sailesh Krishnamurthy
-
Aadhaar Data Vault so simple, even da Vinci would approve. 😐 My latest blog explores how Google Cloud offers a simpler, affordable approach to…
Aadhaar Data Vault so simple, even da Vinci would approve. 😐 My latest blog explores how Google Cloud offers a simpler, affordable approach to…
Liked by Sailesh Krishnamurthy
-
Electronic Arts (EA) has migrated the entire compute capacity of EA SPORTS FC to Google Cloud!
Electronic Arts (EA) has migrated the entire compute capacity of EA SPORTS FC to Google Cloud!
Liked by Sailesh Krishnamurthy
-
research is not my primary activity. however I enjoy publishing as a way to promote knowledge and give back to the community. Apparently, last year…
research is not my primary activity. however I enjoy publishing as a way to promote knowledge and give back to the community. Apparently, last year…
Liked by Sailesh Krishnamurthy
-
Two years ago, we embraced AI and ML in Glance's content management system. Fast forward to today, and we’re transforming nearly 12,000 pieces of…
Two years ago, we embraced AI and ML in Glance's content management system. Fast forward to today, and we’re transforming nearly 12,000 pieces of…
Liked by Sailesh Krishnamurthy
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Sailesh Krishnamurthy
3 others named Sailesh Krishnamurthy are on LinkedIn
See others named Sailesh Krishnamurthy