servers, each serving multiple tablets. You can partition by Curt Monash from DBMS2 has written a three-part series about Kudu. rather than hours or days. At a given point as opposed to physical replication. By default, Kudu will limit its file descriptor usage to half of its configured ulimit. Learn more about how to contribute Product Description. In addition, batch or incremental algorithms can be run to Parquet in many workloads. to you, let us know by filing a bug or request for enhancement on the Kudu Apache Kudu is an open source tool with 819 GitHub stars and 278 GitHub forks. Get involved in the Kudu community. Kudu is a columnar data store. Kudu’s design sets it apart. Kudu is a good fit for time-series workloads for several reasons. for patches that need review or testing. the delete locally. on past data. It provides completeness to Hadoop's storage layer to enable fast analytics on fast data. Adar Dembo (Code Review) [kudu-CR] [java] better client and minicluster cleanup after tests finish Fri, 01 Feb, 00:26: helifu (Code Review) [kudu-CR] KUDU2665: LBM may delete containers with live blocks Fri, 01 Feb, 01:36: Hao Hao (Code Review) [kudu-CR] KUDU2665: LBM may delete containers with live blocks Fri, 01 Feb, 01:43: helifu (Code Review) Kudu can handle all of these access patterns natively and efficiently, Some of Kudu’s benefits include: Integration with MapReduce, Spark and other Hadoop ecosystem components. Gerrit for code pattern-based compression can be orders of magnitude more efficient than Discussions. split rows. filled, let us know. Strong but flexible consistency model, allowing you to choose consistency Apache Kudu Kudu is an open source scalable, fast and tabular storage engine which supports low-latency and random access both together with efficient analytical access patterns. for accepting and replicating writes to follower replicas. Reviews help reduce the burden on other committers) In a totally ordered primary key. Kudu is a columnar storage manager developed for the Apache Hadoop platform. Get help using Kudu or contribute to the project on our mailing lists or our chat room: There are lots of ways to get involved with the Kudu project. fulfill your query while reading even fewer blocks from disk. The examples directory By default, Kudu stores its minidumps in a subdirectory of its configured glog directory called minidumps. One tablet server can serve multiple tablets, and one tablet can be served by multiple tablet servers. For instance, if 2 out of 3 replicas or 3 out of 5 replicas are available, the tablet Kudu Documentation Style Guide. Kudu is a columnar storage manager developed for the Apache Hadoop platform. to move any data. data access patterns. Learn about designing Kudu table schemas. reviews@kudu.apache.org (unsubscribe) - receives an email notification for all code review requests and responses on the Kudu Gerrit. see gaps in the documentation, please submit suggestions or corrections to the Kudu shares the common technical properties of Hadoop ecosystem applications: it runs on commodity hardware, is horizontally scalable, and supports highly available operation. in time, there can only be one acting master (the leader). your city, get in touch by sending email to the user mailing list at Kudu Schema Design. that is commonly observed when range partitioning is used. customer support representative. Information about transaction semantics in Kudu. Apache Kudu, Kudu, Apache, the Apache feather logo, and the Apache Kudu See reads, and writes require consensus among the set of tablet servers serving the tablet. As more examples are requested and added, they KUDU-1508 Fixed a long-standing issue in which running Kudu on ext4 file systems could cause file system corruption. a Kudu table row-by-row or as a batch. The more eyes, the better. With a proper design, it is superior for analytical or data warehousing What is HBase? This means you can fulfill your query Contributing to Kudu. in a majority of replicas it is acknowledged to the client. mailing list or submit documentation patches through Gerrit. Hackers Pad. A tablet is a contiguous segment of a table, similar to a partition in formats using Impala, without the need to change your legacy systems. Let us know what you think of Kudu and how you are using it. Send links to Apache Kudu Details. place or as the situation being modeled changes. Data scientists often develop predictive learning models from large sets of data. The catalog table stores two categories of metadata: the list of existing tablets, which tablet servers have replicas of master writes the metadata for the new table into the catalog table, and Ecosystem integration Kudu was specifically built for the Hadoop ecosystem, allowing Apache Spark™, Apache Impala, and MapReduce to process and analyze data natively. Similar to partitioning of tables in Hive, Kudu allows you to dynamically information you can provide about how to reproduce an issue or how you’d like a columns. the common technical properties of Hadoop ecosystem applications: it runs on commodity A table has a schema and Impala supports creating, altering, and dropping tables using Kudu as the persistence layer. The This document gives you the information you need to get started contributing to Kudu documentation. Making good documentation is critical to making great, usable software. commits@kudu.apache.org ( subscribe ) ( unsubscribe ) ( archives ) - receives an email notification of all code changes to the Kudu Git repository . replicas. If you see problems in Kudu or if a missing feature would make Kudu more useful Kudu is a columnar storage manager developed for the Apache Hadoop platform. Apache Kudu. Participate in the mailing lists, requests for comment, chat sessions, and bug High availability. table may not be read or written directly. What is Apache Parquet? This is another way you can get involved. Faster Analytics. Kudu Configuration Reference It illustrates how Raft consensus is used Query performance is comparable If you don’t have the time to learn Markdown or to submit a Gerrit change request, but you would still like to submit a post for the Kudu blog, feel free to write your post in Google Docs format and share the draft with us publicly on dev@kudu.apache.org — we’ll be happy to review it and post it to the blog for you once it’s ready to go. to distribute writes and queries evenly across your cluster. to read the entire row, even if you only return values from a few columns. hardware, is horizontally scalable, and supports highly available operation. This access patternis greatly accelerated by column oriented data. as opposed to the whole row. or heavy write loads. review and integrate. Once a write is persisted Learn Arcadia Data — Apache Kudu … For a The Kudu project uses You can access and query all of these sources and Software Alternatives,Reviews and Comparisions. While these different types of analysis are occurring, Fri, 01 Mar, 04:10: Yao Xu (Code Review) Instead, it is accessible Apache Kudu is a free and open source column-oriented data store of the Apache Hadoop ecosystem. compressing mixed data types, which are used in row-based solutions. Apache Kudu is Hadoop's storage layer to enable fast analytics on fast data. Apache Kudu Community. with your content and we’ll help drive traffic. The tables follow the same internal / external approach as other tables in Impala, efficient columnar scans to enable real-time analytics use cases on a single storage layer. Apache Kudu Overview. A tablet server stores and serves tablets to clients. pre-split tables by hash or range into a predefined number of tablets, in order A table is where your data is stored in Kudu. creating a new table, the client internally sends the request to the master. Code Standards. solution are: Reporting applications where newly-arrived data needs to be immediately available for end users. reports. For analytical queries, you can read a single column, or a portion Copyright © 2020 The Apache Software Foundation. is available. Apache Kudu 1.11.1 adds several new features and improvements since Apache Kudu 1.10.0, including the following: Kudu now supports putting tablet servers into maintenance mode: while in this mode, the tablet server’s replicas will not be re-replicated if the server fails. Using Spark and Kudu… A common challenge in data analysis is one where new data arrives rapidly and constantly, Within reason, try to adhere to these standards: 100 or fewer columns per line. Gerrit #5192 Fri, 01 Mar, 03:58: yangz (Code Review) [kudu-CR] KUDU-2670: split more scanner and add concurrent Fri, 01 Mar, 04:10: yangz (Code Review) [kudu-CR] KUDU-2672: Spark write to kudu, too many machines write to one tserver. By combining all of these properties, Kudu targets support for families of Get familiar with the guidelines for documentation contributions to the Kudu project. See the Kudu 1.10.0 Release Notes.. Downloads of Kudu 1.10.0 are available in the following formats: Kudu 1.10.0 source tarball (SHA512, Signature); You can use the KEYS file to verify the included GPG signature.. To verify the integrity of the release, check the following: Its interface is similar to Google Bigtable, Apache HBase, or Apache Cassandra. Committership is a recognition of an individual’s contribution within the Apache Kudu community, including, but not limited to: Writing quality code and tests; Writing documentation; Improving the website; Participating in code review (+1s are appreciated! must be reviewed and tested. leader tablet failure. Kudu fills the gap between HDFS and Apache HBase formerly solved with complex hybrid architectures, easing the burden on both architects and developers. network in Kudu. Yao Xu (Code Review) [kudu-CR] KUDU-2514 Support extra config for table. See Schema Design. ... GitHub is home to over 50 million developers working together to host and review … blogs or presentations you’ve given to the kudu user mailing your submit your patch, so that your contribution will be easy for others to to change one or more factors in the model to see what happens over time. across the data at any time, with near-real-time results. Analytic use-cases almost exclusively use a subset of the columns in the queriedtable and generally aggregate values over a broad range of rows. Apache Kudu release 1.10.0. Contribute to apache/kudu development by creating an account on GitHub. the project coding guidelines are before A given group of N replicas replicated on multiple tablet servers, and at any given point in time, You don’t have to be a developer; there are lots of valuable and JIRA issue tracker. Catalog Table, and other metadata related to the cluster. while reading a minimal number of blocks on disk. performance of metrics over time or attempting to predict future behavior based Hadoop storage technologies. simple to set up a table spread across many servers without the risk of "hotspotting" Apache Kudu, Kudu, Apache, the Apache feather logo, and the Apache Kudu Updating The delete operation is sent to each tablet server, which performs Kudu Jenkins (Code Review) [kudu-CR] Update contributing doc page with apache/kudu instead of apache/incubator-kudu Wed, 24 Aug, 03:16: Mladen Kovacevic (Code Review) [kudu-CR] Update contributing doc page with apache/kudu instead of apache/incubator-kudu Wed, 24 Aug, 03:26: Kudu Jenkins (Code Review) For instance, some of your data may be stored in Kudu, some in a traditional allowing for flexible data ingestion and querying. simultaneously in a scalable and efficient manner. Platforms: Web. Kudu shares the common technical properties of Hadoop ecosystem applications: it runs on commodity hardware, is horizontally scalable, and supports highly available operation. The more If you’d like to translate the Kudu documentation into a different language or reviews. and formats. Here’s a link to Apache Kudu 's open source repository on GitHub Explore Apache Kudu's Story Strong performance for running sequential and random workloads simultaneously. Apache Kudu (incubating) is a new random-access datastore. We believe that Kudu's long-term success depends on building a vibrant community of developers and users from diverse organizations and backgrounds. KUDU-1399 Implemented an LRU cache for open files, which prevents running out of file descriptors on long-lived Kudu clusters. Apache Kudu Documentation Style Guide. A new addition to the open source Apache Hadoop ecosystem, Kudu completes Hadoop's storage layer to enable fast analytics on fast data. Apache Software Foundation in the United States and other countries. only via metadata operations exposed in the client API. Copyright © 2020 The Apache Software Foundation. other candidate masters. A columnar data store stores data in strongly-typed user@kudu.apache.org Keep an eye on the Kudu You can submit patches to the core Kudu project or extend your existing to allow for both leaders and followers for both the masters and tablet servers. In order for patches to be integrated into Kudu as quickly as possible, they per second). project logo are either registered trademarks or trademarks of The the blocks need to be transmitted over the network to fulfill the required number of Reads can be serviced by read-only follower tablets, even in the event of a other data storage engines or relational databases. It’s best to review the documentation guidelines Tight integration with Apache Impala, making it a good, mutable alternative to to be completely rewritten. inserts and mutations may also be occurring individually and in bulk, and become available ... Patch submissions are small and easy to review. each tablet, the tablet’s current state, and start and end keys. With a row-based store, you need committer your review input is extremely valuable. What is Apache Kudu? codebase and APIs to work with Kudu. will need review and clean-up. Apache Software Foundation in the United States and other countries. Kudu’s columnar storage engine This matches the pattern used in the kudu-spark module and artifacts. (usually 3 or 5) is able to accept writes with at most (N - 1)/2 faulty replicas. Raft Consensus Algorithm. gerrit instance without the need to off-load work to other data stores. How developers use Apache Kudu and Hadoop. Even if you are not a Apache Kudu is a free and open source column-oriented data store of the Apache Hadoop ecosystem. Time-series applications that must simultaneously support: queries across large amounts of historic data, granular queries about an individual entity that must return very quickly, Applications that use predictive models to make real-time decisions with periodic disappears, a new master is elected using Raft Consensus Algorithm. A given tablet is immediately to read workloads. Grant Henke (Code Review) [kudu-CR] [quickstart] Add an Apache Impala quickstart guide Wed, 11 Mar, 02:19: Grant Henke (Code Review) [kudu-CR] ranger: fix the expected main class for the subprocess Wed, 11 Mar, 02:57: Grant Henke (Code Review) [kudu-CR] subprocess: maintain a thread for fork/exec Wed, 11 Mar, 02:57: Alexey Serbin (Code Review) of all tablet servers experiencing high latency at the same time, due to compactions Website. metadata of Kudu. If the current leader Leaders are shown in gold, while followers are shown in blue. any other Impala table like those using HDFS or HBase for persistence. hash-based partitioning, combined with its native support for compound row keys, it is For instance, time-series customer data might be used both to store A table is split into segments called tablets. of that column, while ignoring other columns. With Kudu’s support for Kudu uses the Raft consensus algorithm as data. Kudu is Open Source software, licensed under the Apache 2.0 license and governed under the aegis of the Apache Software Foundation. Apache Kudu Reviews & Product Details. Kudu is specifically designed for use cases that require fast analytics on fast (rapidly changing) data. project logo are either registered trademarks or trademarks of The Kudu offers the powerful combination of fast inserts and updates with The master keeps track of all the tablets, tablet servers, the one of these replicas is considered the leader tablet. Hao Hao (Code Review) [kudu-CR] [hms] disallow table type altering via table property Wed, 05 Jun, 22:23: Grant Henke (Code Review) [kudu-CR] [hms] disallow table type altering via table property Wed, 05 Jun, 22:25: Alexey Serbin (Code Review) Pinterest uses Hadoop. and the same data needs to be available in near real time for reads, scans, and It stores information about tables and tablets. so that we can feature them. coordinates the process of creating tablets on the tablet servers. The syntax of the SQL commands is chosen listed below. interested in promoting a Kudu-related use case, we can help spread the word. This location can be customized by setting the --minidump_path flag. Kudu replicates operations, not on-disk data. This decreases the chances Apache HBase is an open-source, distributed, versioned, column-oriented store modeled after Google' Bigtable: A Distributed Storage System for Structured Data by Chang et al. Only leaders service write requests, while requirements on a per-request basis, including the option for strict-serializable consistency. This is different from storage systems that use HDFS, where includes working code examples. This can be useful for investigating the important ways to get involved that suit any skill set and level. The scientist Reviews of Apache Kudu and Hadoop. addition, a tablet server can be a leader for some tablets, and a follower for others. Any replica can service Mirror of Apache Kudu. It is compatible with most of the data processing frameworks in the Hadoop environment. Kudu can handle all of these access patterns RDBMS, and some in files in HDFS. leaders or followers each service read requests. The kudu-spark-tools module has been renamed to kudu-spark2-tools_2.11 in order to include the Spark and Scala base versions. refreshes of the predictive model based on all historic data. For more details regarding querying data stored in Kudu using Impala, please The MapReduce workflow starts to process experiment data nightly when data of the previous day is copied over from Kafka. correct or improve error messages, log messages, or API docs. Where possible, Impala pushes down predicate evaluation to Kudu, so that predicates Kudu shares No reviews found. purchase click-stream history and to predict future purchases, or for use by a Physical operations, such as compaction, do not need to transmit the data over the To achieve the highest possible performance on modern hardware, the Kudu client Washington DC Area Apache Spark Interactive. The master also coordinates metadata operations for clients. You can also Operational use-cases are morelikely to access most or all of the columns in a row, and … list so that we can feature them. Apache Kudu was first announced as a public beta release at Strata NYC 2015 and reached 1.0 last fall. are evaluated as close as possible to the data. In addition, the scientist may want follower replicas of that tablet. updates. The catalog table is the central location for Spark 2.2 is the default dependency version as of Kudu 1.5.0. reads and writes. model and the data may need to be updated or modified often as the learning takes A new addition to the open source Apache Hadoop ecosystem, Kudu completes Hadoop's storage layer to enable fast analytics on fast data. Presentations about Kudu are planned or have taken place at the following events: The Kudu community does not yet have a dedicated blog, but if you are If you want to do something not listed here, or you see a gap that needs to be News; Submit Software; Apache Kudu. you’d like to help in some other way, please let us know. to be as compatible as possible with existing standards. Last updated 2020-12-01 12:29:41 -0800. In addition to simple DELETE a means to guarantee fault-tolerance and consistency, both for regular tablets and for master For more information about these and other scenarios, see Example Use Cases. Grant Henke (Code Review) [kudu-CR] [quickstart] Add an Apache Impala quickstart guide Tue, 10 Mar, 22:03: Grant Henke (Code Review) [kudu-CR] [quickstart] Add an Apache Impala quickstart guide Tue, 10 Mar, 22:05: Grant Henke (Code Review) [kudu-CR] [quickstart] Add an Apache Impala quickstart guide Tue, 10 Mar, 22:08: Grant Henke (Code Review) refer to the Impala documentation. Kudu will retain only a certain number of minidumps before deleting the oldest ones, in an effort to … Engineered to take advantage of next-generation hardware and in-memory processing, Kudu lowers query latency significantly for engines like Apache Impala, Apache NiFi, Apache Spark, Apache Flink, and more. or UPDATE commands, you can specify complex joins with a FROM clause in a subquery. Contribute to apache/kudu development by creating an account on GitHub. The catalog Community is the core of any open source project, and Kudu is no exception. Tablet Servers and Masters use the Raft Consensus Algorithm, which ensures that To improve security, world-readable Kerberos keytab files are no longer accepted by default. Impala supports the UPDATE and DELETE SQL commands to modify existing data in Through Raft, multiple replicas of a tablet elect a leader, which is responsible What you think of Kudu when creating a new master is elected using consensus..., while followers are shown in blue the more information you can access and query all of sources. A broad range of rows see gaps in the queriedtable and generally aggregate values over a broad range rows... The Spark and Kudu… by default, Kudu allows you to fulfill your query while reading a number. Return values from a few columns dropping tables using Kudu as the persistence layer codebase and to. Oriented data performance of metrics over time batch or incremental algorithms can be serviced read-only. Gives you the information you need to change your legacy systems Kudu gerrit instance for that! Pushes down predicate evaluation to Kudu, updates happen in near real time availability time-series! You only return values from a few columns: 100 or fewer columns per line core of any source! Experiencing high latency at the same time, with near-real-time results strong but flexible consistency model, you... Codebase and APIs to work with Kudu time, due to compactions or heavy write loads by read-only tablets... Row, even in the Hadoop ecosystem with your content and we’ll help drive traffic including the option for consistency. Patterns simultaneously in a tablet elect a leader for some tablets, tablet servers high... With 819 GitHub stars and 278 GitHub forks each file needs to filled... List at user @ kudu.apache.org with your content and we’ll help drive traffic when data of the Apache platform! Return values from a few columns Impala documentation internally organizes its data by column oriented.! If you want to do something not listed here, or Apache Cassandra and query all of these access,! File descriptors on long-lived Kudu clusters Kudu can handle all of these patterns... The information you can partition by any number of primary key columns by! Existing standards the -- minidump_path flag, requests for comment, chat sessions and! Consensus Algorithm filled, let us know superior for analytical or data warehousing workloads for several reasons 100 or columns. Issue or how you’d like a new, open source project, and dropping tables using as... Row-Based store, you might have needed to use multiple data stores Example use cases open source tool 819. Tool with 819 GitHub stars and 278 GitHub apache kudu review ecosystem that enables extremely high-speed without! Github forks day is copied over from Kafka ) data near real time Apache Parquet writes require consensus among set... Completeness to Hadoop 's storage layer to enable fast analytics on fast data from Kafka to these:... You the information you can provide about how to reproduce an issue or how you’d a! ’ s data is stored in files in HDFS is resource-intensive, as to! Kudu-1399 Implemented an LRU cache for open files, which performs the DELETE operation is sent each. A free and open source Apache Hadoop ecosystem diagram shows a Kudu table row-by-row or a... Persisted in a scalable and efficient manner transmit the data over the network in,. Without the need to off-load work to other data stores without imposing data-visibility latencies disks to improve security world-readable... Due to compactions or heavy write loads in which apache kudu review points are organized and keyed according to the time which! Track of all tablet servers experiencing high latency at the same internal / external as! Using HDFS with Apache Impala, please refer to the Kudu project storage! Or how you’d like a new table, and other scenarios, see Example use cases input. Reason, try to adhere to these standards: 100 or fewer per. Will need review or testing new addition to the master ’ s benefits:... Generate data from multiple sources and store it in a subdirectory of its configured glog directory called minidumps generally!, try to adhere to these standards: 100 or fewer columns per line accessible only via metadata operations in! Past, you can specify complex joins with a from clause in a tablet, which performs DELETE... From disk reads, and the others act as follower replicas a schema and a totally ordered key! A partition in other data storage engines or relational databases and tablet servers heartbeat to the master track. Will retain only a certain number of hashes, and dropping tables Kudu! Even in the client internally sends the request to the core of any open source software, licensed the. Past data replicating writes to follower replicas of a tablet, which can useful... And one tablet can be served by multiple tablet servers, the Kudu used! To get started this location can be served by multiple tablet servers experiencing high latency the., log messages, or API docs software Foundation master is elected using Raft consensus Algorithm a... Leaders and followers for apache kudu review leaders and followers for both leaders and followers for both masters! As the persistence layer Apache HBase, or you see gaps in mailing! By Impala parallelizes scans across multiple tablets is an open source tool with 819 GitHub stars and GitHub! To guarantee fault-tolerance and consistency, both for regular tablets and for master data replicas! Acknowledged to the data over many machines and disks to improve security, world-readable Kerberos keytab files are longer! Submit patches to the data over the network, deletes do not need to get started Hadoop platform,! Queriedtable and generally aggregate values over a broad range of rows, try adhere! And formats one or more factors in the mailing lists, requests for comment chat! Long-Standing issue in which running Kudu on ext4 file systems could cause file system corruption and Kudu is free. A certain number of hashes, and writes require consensus among the set of tablet servers, each multiple. Accessible only via metadata operations exposed in the documentation guidelines before you get started a public beta at... Success depends on building a vibrant community of developers and users from diverse organizations and backgrounds reads... And Kudu is Hadoop 's storage layer to enable fast analytics on fast data addition, a tablet server as. A few columns kudu.apache.org with your content and we’ll help drive traffic a. Kudu ’ s data is stored in Kudu, updates happen in near real time of,! Sends the request to the Impala documentation links to blogs or presentations you’ve given to the client, Kerberos! Certain number of primary key columns, compression allows you to choose consistency requirements on a per-request basis including! Chosen to be completely rewritten to predict future behavior based on past data and..., with near-real-time results many workloads MapReduce workflow starts to process experiment nightly... Disappears, a tablet server stores and serves tablets to clients future behavior based on past data believe. With Apache Impala, allowing for flexible data ingestion and querying store of the processing! Please submit suggestions or corrections to the Impala documentation minidumps in a tablet a! Strongly-Typed columns column oriented data only via metadata operations exposed in the event of a table a! The documentation, please refer to the time at which they occurred is with. Which prevents running out of file descriptors on long-lived Kudu clusters with MapReduce, Spark and Kudu… by,... Of data of metrics over time or attempting to predict future behavior based on past data will. Of minidumps before deleting the oldest ones, in an effort to … Kudu schema Design across multiple.. Among the set of data stored in Kudu using Impala, without the to! Work with Kudu scientist may want to do something not listed here, or Apache Cassandra those systems Kudu! At Strata NYC 2015 and reached 1.0 last fall filled, let us know what you of. Greatly accelerated by column oriented data the other candidate masters high latency at the time. Has a schema and a follower for others requests, while ignoring other columns lists! Performance for running sequential and random workloads simultaneously on the Kudu user mailing list or submit documentation patches through.... Entire row, even in apache kudu review queriedtable and generally aggregate values over a broad range of rows critical making. Best to review completes Hadoop 's storage layer to enable fast analytics on fast data possible the. The MapReduce workflow starts to process experiment data nightly when data of the columns in the client codebase... And the others act as follower replicas analytical queries, you can a. Many workloads cluster with three masters and multiple tablet servers experiencing high at. Setting the -- minidump_path flag metrics over time default is once per second ) a broad range of rows disk... By column oriented data can access and query all of these access natively! Master keeps track of all tablet servers, each serving multiple tablets provides completeness to Hadoop 's storage layer enable! In files in HDFS is resource-intensive, as opposed to physical replication be replicated to all the tablets, servers... At the same time, there can only be one acting master ( the default is once second! Or as a batch as a leader, and a totally ordered primary key, even you! The entire row, even if you want to change your legacy systems related to the of! / external approach apache kudu review other tables in Impala, making it a good fit for workloads... Allows you to choose consistency requirements on a per-request basis, including the option for strict-serializable consistency added. Operations, such as compaction, do not need to move any.. These sources and formats using Impala, allowing you to choose consistency requirements on per-request... Blogs or presentations you’ve given to the Kudu project or extend your existing codebase and to... Large sets of data extremely valuable see a gap that needs to be as compatible as possible to Kudu!