DSE 6.8.x is compatible with Apache Cassandra™ 3.11 and adds additional production-certified changes, if any. Components that are indicated with an asterisk (*) (if any) are known to be updated since the prior patch version.

Release notes of versions prior to 6.8.4 can be found here.

25 April 2024

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.30
• Apache TinkerPop™ 3.4.14-20240307-bcc67d14
• Apache Tomcat® 8.5.94
• DSE Java Driver 1.10.0-dse-20240212 (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Adds two configuration parameters enabling reduction of table histogram metrics cardinality. Those are -Dcassandra.table_metrics_default_histograms_aggregation=[INDIVIDUAL|AGGREGATED] which controls whether tables use individual (default) or keyspace histograms, and -Dtable_metrics_export_globals=[true|false]which controls whether global table histograms exist (default). (DSP-24166)

23 April 2024

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.30
• Apache TinkerPop™ 3.4.14-20240307-bcc67d14
• Apache Tomcat® 8.5.94
• DSE Java Driver 1.10.0-dse-20240212 (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Changed the default location of the histogram aggregation work, offloading it by default to a thread pool different than the main one. This unblocks the TPC threads when large numbers of tables exist. (DSP-24165)

This release is an internal DSE release identical to 6.8.46 release.

12 April 2024

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.30
• Apache TinkerPop™ 3.4.14-20240307-bcc67d14
• Apache Tomcat® 8.5.94
• DSE Java Driver 1.10.0-dse-20240212 (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed millisecond precision point in time restore. (DSP-23993)
• Added support for AWS EC2 IMDSv2. Please beware that if you use Ec2Snitch or Ec2MultiRegionSnitch, by default it will communicate with AWS IMDSv2. This change is transparent and does not need anything done upon upgrade. Consult cassandra-rackdc.properties for more details. (DSP-23995)
• Fixed nodetool viewbuildstatus to query the responsible replica. It prevents returning unknown status when system_distributed RF is smaller than the number of nodes in the cluster. (DSP-23806)

• Fixed Solr credentials parsing. (DSP-24102)

• Added possibility to close/block connection per role by specifying connection_idle_timeout_seconds and connection_idle_behavior via a role custom options. (DSP-23951)

11 March 2024

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.30*
• Apache TinkerPop™ 3.4.14-20240307-bcc67d14*
• Apache Tomcat® 8.5.94
• DSE Java Driver 1.10.0-dse-20240212* (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Reverted the regression caused by DSP-23913 which introduced a change in batch size calculation that impacts the behaviour of the batch_size guardrail. Introduced a new guardrail called batch_size_with_pk_warn_threshold_in_kb, batch_size_with_pk_fail_threshold_in_kb instead that honours the updated logic. (DSP-24011)

• Fixed issue causing indefinite waits during flush operations when TPC executor gets overloaded and default queue size is exceeded. (DSP-23774)
• Modified DSE Advanced Authentication to preserve credentials cache in case of an LDAP internal error causing authentication failure. (DSP-12590)
• Improved LDAP logging by decreasing the frequency of search reference warning messages. (DSP-21177)
• Changed DSE Advanced Authentication to only record in audit log a login error when authentication fails due to matching credentials (and not for provider internal errors). (DSP-23952)

• Upgraded JDK versions in DSE Docker images to 8u392 and 11.0.21. (DSP-23213)

• Upgraded org.json:json to version 20240205. (DSP-23784, CVE-2023-5072)
• Upgraded snappy-java to version 1.1.10.4. (DSP-23819, CVE-2023-43642)
• Upgraded jnr-posix to version 3.1.8. (DSP-23820, CWE-416)

5 February 2024

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.29
• Apache TinkerPop™ 3.4.14-20231030-479dc6d7
• Apache Tomcat® 8.5.94
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed mutation size calculation formula by taking into account static column updates. Backporting CASSANDRA-15293 achieved this fix. (DSP-23933)
• Fixed batch size guardrail to take into account the mutation primary key size. That prevents flooding the cluster with operations for tables that use the primary key without clustering columns. (DSP-23913)

• Fixed DSEFS file path handling that could fail when using filenames containing colons. (DSP-23947)

• Removed Python 2.7 libraries from collectd. (DSP-23764)

• Upgraded the DSE 6.8 dependency on Ehcache to Terracotta's version of Ehcache v2.10.10.17.20. The Terracotta version does not include extra libraries (specifically Jackson databind). The previous Ehcache v2.10.9.2 was exposing a security vulnerability CVE-2020-36518. The vulnerability in jackson-databind before v2.13.0 allowed a Java StackOverflow exception and denial of service via a large depth of nested objects. (DSP-23508, CVE-2020-36518, CVE-2017-17485, CVE-2017-7525, CVE-2018-11307, CVE-2018-7489, CVE-2019-16942)

18 December 2023

• Apache Solr™ 6.0.1.4.2964
• Apache Spark™ 2.4.0.29
• Apache TinkerPop™ 3.4.14-20231030-479dc6d7
• Apache Tomcat® 8.5.94*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.100.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed deadlock in indexes initialization that occurs when the same table has both a secondary index and a search index and the entries in IndexInfo table are missing. The deadlock is resolved by marking the SOLR index as built in a different thread than the main DSE thread. (DSP-23828)

• Fixed the ConcurrentModificationException exception occurring in error during the NodeSync old validations cleanup process. (DSP-23821)

• Upgraded Jetty to version 9.4.53.v20231009. (DSP-23734, CVE-2023-44487)
• Upgraded Apache Tomcat to version 8.5.94. (DSP-23779, CVE-2023-45648)

7 November 2023

• Apache Solr™ 6.0.1.4.2964*
• Apache Spark™ 2.4.0.29
• Apache TinkerPop™ 3.4.14-20231030-479dc6d7*
• Apache Tomcat® 8.5.93
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.100.1.dse*
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Ensured that tombstones get NodeSynced before expiring by assigning segments that have never successfully been NodeSynced an urgent priority. (DSP-23710)

• Upgraded Netty to version 4.1.100.1.dse that is based on 4.1.100.Final. (DSP-23763, CVE-2023-44487, CVE-2022-41881, CVE-2023-34462)
• Removed htrace coming from Hadoop libraries (see HADOOP-17424). Removed jackson-databind version 2.4.0 that was a transitive dependency of htrace. (DSP-23450)
• Removed the htrace version from the lucene-solr library. htrace is an unused dependency in DSE 6.8. This removal resolved security vulnerabilities related to the htrace dependency, despite its being unused. (DSP-23756)

9 October 2023

• Apache Solr™ 6.0.1.4.2959
• Apache Spark™ 2.4.0.29
• Apache TinkerPop™ 3.4.14-20230814-301fd418
• Apache Tomcat® 8.5.93
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Ensured that DSE uses only version 1.12.x of the aws-sdk-java library. Removed the dependency on version 1.11.x which also eliminated the need for the outdated and vulnerable jackson-databind version 2.6.7.3. (DSP-23613)

11 September 2023

• Apache Solr™ 6.0.1.4.2959
• Apache Spark™ 2.4.0.29*
• Apache TinkerPop™ 3.4.14-20230814-301fd418*
• Apache Tomcat® 8.5.93*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Changed logging level from error to warn for a log message that is issued when folders are removed during a snapshot that calculates folder size. (DSP-23432)
• Added a check to prevent running cleanup operations concurrently with operations that lead to token ownership changes, such as node addition or node decommission. Prior to this fix, running such concurrent operations could unintentionally delete valid data replicas. (DSP-23507)

• Changed reads from an SSTable to be delimited by and not exceed the declared row size in the row preamble. This prevents Out of Memory issues (OOM) with a corrupted SSTable. (DSP-23336)

• Upgraded snappy-java to version 1.1.10.3. (DSP-23499)

• Improved SAI queries response time by reducing the number of skips in the predicate intersection algorithm when selectivity of predicates varies significantly. (DSP-23435)
• Added a JVM configuration option to disable Storage Attached Index (SAI) segment compaction. Disable compaction by setting the cassandra.sai.enable_segment_compaction JVM flag to false. The default value is true. (DSP-23440)
• Fixed SAI index build failure for huge SSTables. (DSP-23478)

• Changed to use collectd v0.1.6 bundle based on Ubuntu:18.04. (DSP-23519)

• Fixed the nodetool repair --trace command to prevent it from hanging when it is run on an empty keyspace or on a keyspace with nodesync-enabled tables. (DSP-23408)

• Fixed a bug where Key Management Interoperability Protocol (KMIP) server failover was not working as intended because of exceptions that changed in the KMIP client library. (DSP-23343)

• Upgraded SnakeYAML library to the latest 2.0 version. (DSP-23429, CVE-2022-1471)
• Upgraded java-xmlbuilder to version 1.3. (DSP-23489, CVE-2014-125087)
• Upgraded Apache Tomcat to version 8.5.93. (DSP-23522, CVE-2023-41080)
• Upgraded ‘Google Guava’ to version 32.1.2-jre to remove CVE-2023-2976. Upgraded ‘FasterXML Jackson’ libraries to version 2.13.5. (DSP-23525, CVE-2023-2976)
• Enforced net.sf.ehcache library to use version 2.10.9.2 instead of 2.10.4. Removed indirect dependency on jackson-databind version 2.3.3. (DSP-23528)

10 July 2023

• Added support for Red Hat Enterprise Linux 9. (DSP-23229)
• Added support for Oracle Linux 9. (DSP-22612)

• Apache Solr™ 6.0.1.4.2959
• Apache Spark™ 2.4.0.28
• Apache TinkerPop™ 3.4.14-20230523-37856751
• Apache Tomcat® 8.5.89
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.54

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Allowed to encrypt kmip_host passwords in the dse.yaml file. New optional parameter password_encryption_key_name was added to the kmip_hosts items. DSP-23326 enabled password encryption before placing them in dse.yaml. (DSP-23313)
• Fixed the issue with dropping the UDT column, which potentially leads to a problem accessing data from the affected table after the node is restarted. (DSP-23337)

• Fixed race condition causing sstableloader to hang when the Connections Per Host (-cph) value is greater than one (1). (DSP-23117)

• Updated official DSE Docker image, basing it on a more current Ubuntu 20.04 (Focal Fossa) that is FIPS certified and handles multiple known CVEs. (DSP-23397)

• Improved SAI index-building process when reloading SSTables by building the indexes in parallel. (DSP-23415)
• Reduced allocations by removing some unnecessary autoboxing from SAI range iterators. (DSP-23419)
• Improved SAI exact match queries performance on the memtable index data by skipping unnecessary size calculations. (DSP-23280)
• Fixed a bug in the SAI intersection comparison. (DSP-23375)

• Upgraded Apache mina-core library to version 2.0.24. (DSP-23378, CVE-2021-41973)
• Upgraded Azure SDK client libraries to be based on BOM 1.2.13. (DSP-23382, CVE-2023-1370)
• Updated aws-java-sdk to version 1.12.486. (DSP-23383, CVE-2022-31159)
• Upgraded Apache Commons libraries to recently available versions: commons-io to 2.11.0, commons-lang3 to 3.12.0, commons-math3 to 3.6.1, and commons-collection4 to 4.4. (DSP-23384, CVE-2021-29425)
• Upgraded xerial/snappy-java library to version 1.1.10.1. (DSP-23391, DSP-23433, CVE-2023-34453, CVE-2023-34454, CVE-2023-34455)
• Upgraded Jetty to version 9.4.51.v20230217. (DSP-23428, CVE-2023-26048)
• Removed all dependencies on Netty 3 known to have security vulnerabilities. (DSP-22562, CVE-2019-16869)

12 June 2023

• Apache Solr™ 6.0.1.4.2959
• Apache Spark™ 2.4.0.28*
• Apache TinkerPop™ 3.4.14-20230523-37856751*
• Apache Tomcat® 8.5.89*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.54*

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed compaction to ensure tombstones are not pruned for data in SSTables that are excluded from the compaction due to limited disk space. This fixes CASSANDRA-18507. (DSP-23305)

• Improved logging around index summary building. Introduced rounding that effectively uses the min_index_interval option set to a multiple of 128 in the index summary building process. (DSP-23317)

• Fixed the bug in the Solr UI permissions when accessing segments information. It was not working with authorization enabled by implementing a segments view when granting the SELECT permission on the Solr-indexed table. (DSP-22295)

• Added --key option to dsetool encryptconfigvalue command to allow using different key than the default one. (DSP-23326)

• Fixed memory leak in the chunk cache metadata caching. (DSP-23316)

• Upgraded commons-net library used in Hadoop and Tinkerpop. Removed old Hadoop 1.0.3 dependency and replaced it with version universally used in DSE (2.7.1.x for 5.1 and 2.10.2.x for 6.8). (DSP-23327, CVE-2021-37533, CVE-2012-4449)
• Upgraded Apache Tomcat to version 8.5.89. (DSP-23329, CVE-2023-28709)

10 May 2023

• Apache Solr™ 6.0.1.4.2959
• Apache Spark™ 2.4.0.27
• Apache TinkerPop™ 3.4.14-20230215-3db1ca33
• Apache Tomcat® 8.5.87
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.53

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed regression in KMIP encryption introduced in DSP-22164. Moved the logic required for using old HashiVault KMIP under the feature flag experimental_custom_attributes_mode in dse.yaml, defined per kmip_hosts.per-kmip-group section. (DSP-23278)

• Fixed SAI numeric index validation on SSTables with over 2 billion rows. (DSP-23255)

• Improved CQLSH to support Python 3.11 and changed preferred interpreter from python2 to python3. Introduced CQLSH_PYTHON environment variable to allow users to specify the python interpreter used by CQLSH. (DSP-23257)

• Upgraded org.codehaus.jettison:jettison library to version 1.5.4. (DSP-23254, CVE-2023-1436)
• Upgraded com.google.auto.factory:auto-factory to version 1.0.1 to remove the dependency on org.eclipse.equinox. (DSP-22864, CVE-2021-41033)

11 April 2023

• Apache Solr™ 6.0.1.4.2959*
• Apache Spark™ 2.4.0.27
• Apache TinkerPop™ 3.4.14-20230215-3db1ca33
• Apache Tomcat® 8.5.87*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.53

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Set default systemd dse.service startup timeout to 7 minutes. (DSP-23225)

• Fixed memory leak that can happen while handling inbound large messages. (DSP-23156)

• Fixed race condition occurring in DSE Search with enabled real-time (RT) indexing that was causing client-side failures under query-heavy workload. (DSP-23184)

• Updated OS versions used to create DSE packages (Ubuntu 12.04 for .deb packages, and CentOS 7.9 for .rpm packages). (DSP-22750)

• Fixed regression scenario where DSE was not using keys on the KMIP server that were created either by a previous DSE version or outside of DSE. Regression was introduced in DSE v6.8.22. (DSP-23182)

• Upgraded org.json:json to 20230227 to resolve a Denial of Service (DoS) vulnerability. In addition, upgraded esri-geometry-api to 2.2.4. (DSP-23187, CWE-400)
• Upgraded commons-fileupload to 1.5. Added a solrconfig.xml setting that limits the number of files allowed in multipart update requests. (DSP-23188, CVE-2023-24998)
• Upgraded Apache Tomcat to version 8.5.87. (DSP-23205, CVE-2023-24998)

6 March 2023

• Apache Solr™ 6.0.1.4.2951*
• Apache Spark™ 2.4.0.27*
• Apache TinkerPop™ 3.4.14-20230215-3db1ca33*
• Apache Tomcat® 8.5.84
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.53*

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Exposed two configuration properties: -Dcassandra.counter_locks_per_core and -Dcassandra.lwt_locks_per_core to be tuned for intensive counter batch workloads when observing WriteTimeoutExceptions. For example, update the default of 1024 to -Dcassandra.lwt_locks_per_core=16384. (DSP-23163)
• Fixed SAI index metrics initialization potentially causing dse start failure due to NullPointerException. (DSP-23192)

• Added support for Rocky Linux 9. (DSP-22732)
• Added support for Rocky Linux 8. (DSP-23170)
• Added support for Oracle Linux 9. (DSP-22612)

• Fixed RPM package install for platforms requiring systemd services. (DSP-23146)

• Upgraded libthrift to v0.9.3-1. (DSP-18096, CVE-2018-1320)
• Upgraded groovy-sandbox to 1.20.1.DSE, which is a DataStax version based on OSS v1.20 that contains additional fixes. (DSP-21677, CVE-2018-1000865)
• Ported a security fix from Spark 2.4.6 that prevents RCE on unauthenticated Spark resource manager. (DSP-21782, CVE-2020-9480)
• Removed Postgresql driver from spark-jobserver. (DSP-22894, CVE-2022-21724, CVE-2020-13692, CVE-2018-10936)
• Upgraded insights-collectd to version 0.1.5 that removed libmodbus.so. (DSP-22809, CVE-2022-0367)
• Upgraded Gson that is used in Solr to v2.10.1. (DSP-22798, CVE-2022-25647)
• Upgraded Apache Derby used in Spark to v10.14.2.0. (DSP-23008, CVE-2018-1313)
• Removed unused, outdated org.mortbay.jetty libraries. (DSP-23004, CVE-2011-4461)
• Upgraded Jettison to 1.5.3, Xerces to 2.12.2, and Gson Hadoop to 2.10.1. Enforced the same Hadoop 2.10.2.x version for all DSE components. (DSP-23120, CVE-2022-40149, CVE-2022-40150, CVE-2013-4002)
• Upgraded Jcommander used in SJK to v1.82. (DSP-21783, SRCCLR-SID-22555)

7 February 2023

• Apache Solr™ 6.0.1.4.2947*
• Apache Spark™ 2.4.0.24*
• Apache TinkerPop™ 3.4.14-20230124-b2e47e9a*
• Apache Tomcat® 8.5.84
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse
• Spark JobServer 0.8.0.52

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Added logging of unexpected failures in asynchronous reactive code. (DSP-23111)

• Fixed incorrect count(*) results when hitting guardrail paging limits. (DSP-22813)

• Allow the use of date with time in Solr range queries on SimpleDateField fields (time will be truncated). (DSP-23085)

• Fixed the CLI usage of namespace filtering in the dsetool utility. (DSP-23064)
• Fixed the dsetool utility ‘managekmip revoke’ command to correctly set the ‘compromiseOccurrenceDate’ field. (DSP-23110)

• Fixed error preventing cqlsh COPY from Mac M1 hardware. (DSP-22996)

• Upgraded httpclient version to 4.5.14 version. (DSP-22831, CVE-2020-13956)
• Upgraded Spark to version that is based on Hadoop 2.10.2. (DSP-22923, CVE-2016-6811, CVE-2017-3162)
• Upgraded TinkerPop to a version that uses Hadoop 2.10.2. (DSP-23092, CVE-2022-25168, CVE-2021-33036, CVE-2021-37404, CVE-2020-9492, CVE-2018-8009, CVE-2016-3086, CVE-2016-6811, CVE-2016-5393)

9 January 2023

• Apache Solr™ 6.0.1.4.2943
• Apache Spark™ 2.4.0.23
• Apache TinkerPop™ 3.4.14-20221125-fd3c10f9
• Apache Tomcat® 8.5.84*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.86.1.dse*
• Spark JobServer 0.8.0.52*

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Added a guarding exception that prevents the creation of an invalid range tombstone. Improved range tombstone addition by skipping the creation of a tombstone for the empty range. (DSP-23075)

• Upgraded Hadoop direct dependency to 2.10.2.1, a new version that is based on OSS 2.10.2. (DSP-22922)

• Upgraded apache-shiro used by spark-jobserver to version 1.10.1. (DSP-23019, CVE-2022-32532, CVE-2022-40664)
• Upgraded Netty to version 4.1.86.Final. (DSP-23062, CVE-2022-41915, CVE-2022-41881)
• Upgraded Tomcat to version 8.5.84. (DSP-23017, CVE-2022-34305)

12 December 2022

• Apache Solr™ 6.0.1.4.2943*
• Apache Spark™ 2.4.0.23*
• Apache TinkerPop™ 3.4.14-20221125-fd3c10f9*
• Apache Tomcat® 8.5.79
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.78.1.dse
• Spark JobServer 0.8.0.51

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Changed log messages added when Insights functionality cannot be started from ERROR to WARN. (DSP-22962)

• Fixed Debian package dependencies on Python so that it can be installed on Ubuntu 22.04. (DSP-22961)

• Use com.google.cloud:libraries-bom:26.1.3 libraries for GCE backup and restore functionality. Upgraded Guava from 19.0 to 31.1-jre. (DSP-22904, CVE-2020-8908, CVE-2018-10237, CVE-2020-7692)
• Updated Apache Ivy of DSE Spark to version 2.5.1. (DSP-22949, CVE-2022-37865, CVE-2022-37866)
• Removed lucene-benchmark from lucene-solr as it contained unnecessary vulnerable library: nekohtml:1.9.17. (DSP-22902, CVE-2022-28366, CVE-2022-24839, CVE-2022-29546)

14 November 2022

• Apache Solr™ 6.0.1.4.2940
• Apache Spark™ 2.4.0.21*
• Apache TinkerPop™ 3.4.5-20220728-e115ab9a
• Apache Tomcat® 8.5.79
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.78.1.dse
• Spark JobServer 0.8.0.51

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed an issue in DSE Spark which was unable to read S3 objects with $ and = characters in the path. Updated aws-java-sdk to 1.11.892 and hadoop to 2.7.1.5. Dropped support of s3n, which is no longer undergoing active maintenance from the OSS Hadoop. (https://hadoop.apache.org/docs/current2/hadoop-aws/tools/hadoop-aws/index.html#S3N) Users are recommended to migrate to s3a instead. (DSP-22737)

• Upgraded jackson-databind to version 2.13.4.2. (DSP-22905, CVE-2022-42003)
• Replaced the vulnerable woodstox-core library used by jackson-dataformat-xml with version 6.4.0. (DSP-22914, CVE-2022-40151, CVE-2022-40152, CVE-2022-40153, CVE-2022-40154, CVE-2022-40155, CVE-2022-40156)

28 October 2022

• Apache Solr™ 6.0.1.4.2940
• Apache Spark™ 2.4.0.20*
• Apache TinkerPop™ 3.4.5-20220728-e115ab9a
• Apache Tomcat® 8.5.79*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.78.1.dse*
• Spark JobServer 0.8.0.51

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed regression in management api in 6.8.27 by ensuring that epoll netty libraries are included. (DSP-22895)

25 October 2022

NOTE: DSE 6.8.27 has a regression that hinders the use of the management api used in kubernetes. Use of DSE 6.8.27 is not adviced. DSE 6.8.28 resolves this regression.

• Apache Solr™ 6.0.1.4.2940
• Apache Spark™ 2.4.0.20*
• Apache TinkerPop™ 3.4.5-20220728-e115ab9a
• Apache Tomcat® 8.5.79*
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.78.1.dse*
• Spark JobServer 0.8.0.51

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed nodetool flush assertion failure on TWCS tables with split_during_flush=true. (DSP-20708)

• Fixed dsetool core_indexing_status to show consistent indexing status between calls with and without --all parameter. (DSP-21594)

• Fixed GraphOLAP spark connection problem in the multi-network cloud environments when GossipingPropertyFileSnitch (GPFS) is not used. (DSP-22707)
• Fixed permission issue in classic graph where DROP permissions were erroneously required for creating and altering a graph schema. (DSP-22024)

• Upgraded insightsCollectd to 0.1.4. (DSP-22739)

• Fixed an issue where EMPTY_LAST_CONTENT would no be written to the stream when the connection was closed. (DSP-22671)

Fixed an issue during DSE Spark cluster upgrade where InClusterAuthenticator would fail to compose and decode a token if it receives an old version token. (DSP-22723)

• Upgraded jackson-databind to 2.13.4. (DSP-22780, CVE-2022-42004)
• Upgraded Netty to version 4.1.78. (DSP-22511, CVE-2019-9512, CVE-2019-9514, CVE-2019-9515, CVE-2019-20444, CVE-2019-20445, CVE-2020-7238, CVE-2020-11612, CVE-2021-37136, CVE-2021-37137)
• Upgraded version of Apache Tomcat from 8.5.75 to 8.5.79. (DSP-22746, CVE-2022-34305, CVE-2022-29885)
• Upgraded org.apache.commons:commons-text to version 1.10.0. (DSP-22816, CVE-2022-42889)
• Upgraded SnakeYAML to 1.33. (DSP-22773, CVE-2022-25857)
• Upgraded jetty to 9.4.49.v20220914. (DSP-22774, CVE-2022-2048)

12 September 2022

• Apache Solr™ 6.0.1.4.2940
• Apache Spark™ 2.4.0.19
• Apache TinkerPop™ 3.4.5-20220728-e115ab9a*
• Apache Tomcat® 8.5.75
• DSE Java Driver 1.10.0-dse-20220616 (DSE internal-only version)
• Netty 4.1.34.3.dse
• Spark JobServer 0.8.0.51

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed a bug in streaming that could connect back using the wrong address under a specific network setup. (DSP-22585)

• Ported CASSANDRA-15271 to make CLUSTERING ORDER validation less strict. (DSP-21801)
• Fixed ArrayIndexOutOfBoundsException during indexing SAI with ascii option enabled. (DSP-22601)

• Fixed the logic that orders columns in the SELECT list to maintain the given order of the column headers. (DSP-22420)

• Fixed command injection vulnerability in management-api. (DSP-22272)

• Made ‘nodetool status’ report the state of unreachable nodes in order to allow knowing why they are unreachable. (DSP-22648)

18 July 2022

• Apache Solr™ 6.0.1.4.2940
• Apache Spark™ 2.4.0.19*
• Apache TinkerPop™ 3.4.5-20220405-a52bbe2c
• Apache Tomcat® 8.5.75
• DSE Java Driver 1.10.0-dse-20220616* (DSE internal-only version)
• Netty 4.1.34.3.dse
• Spark JobServer 0.8.0.51*

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Upgraded internally used DSE Java driver to 1.10.0-dse-20220616. (DSP-22380) NOTE:
• Replaced 1.9.13 version of jackson-core-asl and jackson-mapper-asl to be 1.9.13.1.dse which contains the security fix. (DSP-22389, CVE-2019-10172)
• Removed the PMD plugin from gradle config. (DSP-22575, CVE-2019-7722)
• Fixed the cqlsh tool to restore the ability to use the EXECUTE AS command. (DSP-22417)
• Fixed bug with overriding customer limits.d file if there are new values in the package upgrade. Complementary to DSP-21928. (DSP-22594)

• Changed outdated nofile parameter value to 1048576 in /etc/security/limits.d/cassandra.conf. (DSP-21947)

• Fixed parsing range queries on DSE Search indexed columns of CQL types Date and Time. (DSP-22548)
• Added SolrCore index size metric. (DSP-22546)
• Added SolrCore numDocs, maxDoc, and deletedDocs metrics. (DSP-22587)
• Added per core DSE Search indexing status as a metric. (DSP-22592)
• Changed Tomcat {{showServerInfo}} configuration parameter to {{false}} for not exposing Tomcat version in error pages. (DSP-22561)

• Upgraded jackson and jackson-databind to 2.13.3. (DSP-22452, CVE-2020-36518)
• Upgraded Spark library version containing change to sanitize passwords when printing spark commands to stderr. (DSP-22481)

• Fixed an error when using the search index and traversing with an inequality operator such as gt or gte against Date or Time data types. (DSP-21279)

• Improved Management API logic to reduce the likelihood of resource leaks when DSE is starting up. (DSP-22539)

• Fixed EXECUTE AS functionality to work with RLAC. (DSP-22508)

• Upgraded apache-shiro used by spark-jobserver to version 1.8.0. (DSP-22557, CVE-2021-41303)

15 June 2022

• Apache Solr™ 6.0.1.4.2940*
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20220405-a52bbe2c
• Apache Tomcat® 8.5.75
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.34.3.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Implemented ability to replace the THREE consistency level with an ALL_BUT_ONE consistency level. (DSP-22366)
• Removed Netty 3.6.2 and 3.7.0. (DSP-22381, CVE-2015-2156)
• Upgraded Jetty All Core to latest version (9.4.46.v20220331). (DSP-22491, CVE-2009-4611, CVE-2020-27216)
• Upgraded Azure SDK client libraries to be based on BOM 1.2.2 to remove CVE-2020-36518 and CVE-2020-5404. (DSP-22528, DSP-21781)
• Fixed the TIMING feature in cqlsh which throws a message "global name 'request_start' is not defined" when enabled. (DSP-22435)

• Added two configurable variables stream_outbound_permits_in_mb and input_stream_channel_timeout_in_ms to allow streaming of high density of data by using ZCS. (DSP-22362)
• Fixed releasing of resources (including heap memory) for metrics removed from the registry (e.g. after table is dropped). (DSP-22516)
• Ported CASSANDRA-16987 and fixed python version check bug in cqlsh. (DSP-22517)

• Upgraded solr version which uses upgraded metadata-extractor 2.18.0 and XmpCore 6.1.11 libraries. (DSP-22406, CVE-2019-14262)

• Fixed dsetool expirekey/revoke commands by using KMIP custom attributes. (DSP-22421)

11 May 2022

• Apache Solr™ 6.0.1.4.2919*
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20220405-a52bbe2c
• Apache Tomcat® 8.5.75
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.34.3.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Improved reading logic to ensure that sstables are not unnecessarily read for columns that are not selected. See CASSANDRA-16737. (Previously DB-4974). (DSP-22478)
• Fixed the URISyntaxException: Malformed IPv6 address when using nodetool or dsetool with Java 8u331 or 11.0.15. This is due to the recent changes of JDK-8278972, in which parsing of URL Strings in Built-in JNDI Providers is more strict. (DSP-22474)

• Greater than '>' and less than '<' operators are swapped in the slow query log for a table with DESC clustering keys (port CASSANDRA-15503). (DSP-22369)
• Fixed a rare race condition where attempting to read from a sstable would fail with an assertion error. (DSP-22431)

• Upgraded xmlbeans version to 4.0.0. (DSP-22379, CVE-2021-23926)
• Upgraded Rome Library to 1.17.0 that uses JDOM 2.0.6.1 version. (DSP-22405, CVE-2021-33813)

• Do not skip 5.1.x format sstables when running nodetool upgradesstables without --include-all-sstables. (DSP-22424)

11 April 2022

• Introduced DSE support for Java 11 for core Cassandra workloads. Please note that DSE does not currently provide support for advanced workloads (Search, Spark and Graph) for Java 11.

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20220405-a52bbe2c*
• Apache Tomcat® 8.5.75*
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.34.3.dse*
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Added a startup check that prevents starting DSE with advanced workloads on Java 11. (DSP-22358)
• Upgraded netty to 4.1.34 from 4.1.25. (DSP-22363)
• Removed disk_cal.py and compaction-metrics.ipynb Python 2 tools. (DSP-22382)
• Used azure sdk client libraries from bill of materials 1.2.0 for Azure Blob Storage backup and restore. Used okhttp based azure code http client instead of the netty based one. (DSP-22401)
• Reduced node bootstrapping and rebuild operation time by improving the algorithm that calculates the map of streaming candidates. (DSP-22339)

• Limited heap pressure during mutation repair for tables with materialized views by throttling number of concurrent batches (default 10). Number of batches can be controlled by new system property cassandra.repair.mutation_repair_max_concurrent_batches. Setting to 0 (zero) disables throttling and reverts behavior before this change. (DSP-22344)

• Replaced log4j with reload4j in TinkerPop and bumped the version of TinkerPop. (DSP-22326)

• Changed heap size to 512M for command line tools on Azul Zing if MAX_HEAP_SIZE is not specified. (DSP-22313)

• Upgraded Bouncy Castle to the latest 1.70 version. (DSP-22352)

• Ported fix from DSP-22315: Option to disable call to NativeLibrary.trySkipCache. (DSP-22343)

• Upgraded azure-storage-blob from 12.4.0 to 12.15.0 version. (DSP-22377, CVE-2020-5403)
• Upgraded apache-commons compress library to 1.21 version. (DSP-22383, CVE-2021-35515, CVE-2021-35516, CVE-2021-35517, CVE-2021-36090)
• Upgraded snakeyaml version to 1.30. (DSP-22386, CVE-2017-18640)
• Upgraded ApacheVelocity to 2.3 version. (DSP-22387, CVE-2020-13936)
• Upgraded commons-beanutils version to 1.9.4 version. (DSP-22388, CVE-2019-10086)
• Upgraded Hazelcast to 5.1.1 version. (DSP-22390, CVE-2022-0265)
• Upgraded logback version to 1.2.11. (DSP-22237, CVE-2021-42550)
• Upgraded version of Apache Tomcat from 8.5.72 to 8.5.75. (DSP-22360, CVE-2022-23181)
• Upgraded version of azure-identity from 1.1.0 to 1.4.6. (DSP-22194, CVE-2017-1000190)

7 March 2022

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2
• Apache Tomcat® 8.5.72
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Changes reading logic of compressed chunk offsets that are loaded from compression info for zero copied partial sstables. It results in smaller off-heap usage. (DSP-22247)
• Adds configurable snapshot size cache speeding up the retrieval of snapshot information. (DSP-22338)
• PKCS#11 needs Signature algorithms to be configured on some versions of Java 11 (see [https://bugs.openjdk.java.net/browse/JDK-8217611|https://bugs.openjdk.java.net/browse/JDK-8217611|smart-link]). In order to work with TLSv1.3, RSA keys in PKCS#11 key stores must have a key size of at least 4096 bits. (DSP-22276)
• RSA Certificates for SSL on Java 11 require larger keypairs than on Java 8. Testing on Java 11 is done with 1024 bit keys instead of 512 bit keys. (DSP-22277)
• Clean up ClientWarn State when message sending expired. (DSP-22290)

• fix deletion of a dropped vertex’s incoming edges when the far side of those edges involves multiple vertex labels. (DSP-22218)

• Fix validation age in the belated incremental NodeSync log warning. (DSP-22300)

• Removed log4j 1.2.x dependency from dse-spark/client/lib and replace it with reload4j 1.2.19. (DSP-22279, CVE-2021-44228, CVE-2019-17571, CVE-2022-23305, CVE-2022-23302, CVE-2021-4104)
• Upgraded version of Bouncy Castle to 1.67. (DSP-22301, CVE-2018-1000613, CVE-2018-1000180, CVE-2020-28052)

17 February 2022

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2
• Apache Tomcat® 8.5.72
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Ported fix from CASSANDRA-17352: Remote code execution for scripted UDFs (DSP-22321, CVE-2021-44521)

24 January 2022

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2
• Apache Tomcat® 8.5.72
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fix a possible overflow with elapsed nano-time calculation in messaging queue timeout. (DSP-22011)

• Await timeout for shutting down non periodic tasks is now configurable with the new jvm option cassandra.non_periodic_tasks_shutdown_timeout_in_minutes. When timeout is reached, force shutdown those tasks. (DSP-22241)
• Lower commitlog replay sstable origin warning to info. (DSP-22270)
• Fix the miscalculation of totalCDCSizeOnDisk. (DSP-22135)

• Retain changes to /etc/security/limits.d/cassandra.conf on yum upgrade. (DSP-21928)

• Removed unused log4j dependency to avoid false positives in vulnerability scans. (DSP-22234, CVE-2019-17571, CVE-2021-4104, CVE-2021-44228, CVE-2021-45105)
• Upgraded version of json-smart library used by Azure blob store access from 2.3 to 2.4.7 to fix CVE-2021-27568 and CVE-2021-31684. (DSP-22186, CVE-2021-27568, CVE-2021-31684)

8 December 2021

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2
• Apache Tomcat® 8.5.72*
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fix calculation of CompressionMetadataOffHeapMemoryUsed metric accessible via nodetool tablestats command. (DSP-22181)

• Port and adjust CASSANDRA-16686 for DSE. (DB-5022)

• Upgraded version of Apache Tomcat from 8.5.70 to 8.5.72 to fix CVE-2021-42340. (DSP-22098)

• Fixed and updated javax.mail dependency to com.sun.mail. (DSP-22085)

• Upgraded version of Apache Tomcat from 8.5.70 to 8.5.72 to fix CVE-2021-42340. (DSP-22098)

4 November 2021

• Apache Solr™ 6.0.1.4.2887
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2
• Apache Tomcat® 8.5.70
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Enables periodic logging of system status (default every 5 minutes, configurable). (DSP-22039)
• Fix a possible issue that shell tool could break with -h/--help in package install. (DSP-20375)

• Fixed a bug in the authenticator that would use the default management mode instead of the defined mode by authentication when authenticating. (DSP-22067)
• Fixes stack overflow with secondary indexes on collections. (DSP-22070)
• Fixed an issue in preloading prepared statements that queries static columns. (DB-5012)

• Fixes issue that when disabling system_info_encryption configuration in dse.yaml, encryption for tables in the system_backups keyspace would not turn off. (DSP-22078)

• Prints TLS protocol information when running cqlsh with --debug parameter. (DB-4981)

• Fixes broken partition filtering in hive metastore leading to missing data in the spark-sql queries results for queries involving numeric partition keys or complex conditions. (DSP-21651)

7 October 2021

• Apache Solr™ 6.0.1.4.2887*
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20210816-c28c0de2*
• Apache Tomcat® 8.5.70*
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• The Change Data Capture (CDC) is now near real time as in Cassandra 4.x. Active commitlog segments can be processed from the CDC raw directory defined in your cassandra.yaml without flushing tables. It also allows you to read the last mutation form the most recent commitlog file. See the Datastax Change Data Capture documentation page for more details. (DSP-21992)

• Added check against the negative value in local stream throughput stream_throughput_outbound_megabits_per_sec and inter dc stream throughput inter_dc_stream_throughput_outbound_megabits_per_sec (DB-5010)
• Fixes output of compaction file progress. (DB-5028)

• Resolves a TPC weakness with large rows and collections, where DSE 6 would repeatedly attempt to read the same row and create a lot of on-heap garbage. (DB-3962)

• Ported fix from SOLR-12514 to dse lucene-Solr to fix CVE-2018-11802. (DSP-21685)
• Upgraded version of PDFBox and FontBox to 2.0.24, and version of JempBox to 1.8.16 to fix CVE-2018-8036 and CVE-2018-11797. (DSP-21688)
• Upgraded version of groovy to 2.4.21 (DSE 5.1/6.0/67) and to 2.5.14 (DSE 6.8) to fix CVE-2020-17521. (DSP-21767)
• Upgraded version of Tomcat from 8.5.65 to 8.5.70 to fix CVE-2021-33037. (DSP-21996)

31 August 2021

• Provide DSE support for Centos8, Red Hat Enterprise Linux 8 and Oracle Linux (DSP-19104). Please note that certification was done using Python 2.7 and Python 2.7 needs to be available on the target system. SparkR version included in DSE 6.8.15 is not compatible with the default R version >= 4.0.0.

26 August 2021

• Apache Solr™ 6.0.1.4.2840
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.5.65
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• upgraded Bootstrap version from 3.1.1 to 3.4.1, upgraded Flask from 0.10.1 to 1.1.4. (DSP-21682)
• Upgraded version of directory-ldap-api from DSE 1.0.0.2.dse to OSS 1.0.3 (DSP-21758)

• Spark had an unused dependency on azure-storage-blob jar files version 12.4.0 which was removed from Azure repos. This fix removes the dependency. (DSP-21978)

Execution of the back pressure task can be rejected in TPC, leading to the back pressure job being dropped entirely and then deadlocking. This was fixed by properly rescheduling the task. (DB-5027)

07 July 2021

• Provide DSE support for Ubuntu 20.04 (Focal) (DSP-21330). Please note that certification was done using Python 2.7 and Python 2.7 needs to be available on the target system.

• Apache Solr™ 6.0.1.4.2840
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.5.65
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Added warning message in case of dse start failure due to this issue (DSP-21795)

• Fixed concurrent modification exception in consistent replace (DSP-21836)

• Upgraded version of resteasy to 4.6.0.Final (DSP-21683)
• Upgraded jetty version from 9.4.34.v20201102 to 9.4.41.v20210516 (DSP-21684, DSP-21687)

• Fixed a bug where in rare cases search query routing might start to spin endlessly for a particular query (DSP-21838)

• Fixed an issue in the LDAP group_search_filter default value that meant that group hierarchies were not being loaded if the group_search_filter was not explicitly set in the dse.yaml. (DSP-21874)

• Removed a possible false-positive error message in the log that would cause confusion when multiple authentication schemes are defined. (DB-5015)

• Added unit Testcases for logic cqlsh TLS version. (DB-4979)

• When the Bloom filter is recreated due to FP chance change, sstable metadata is loaded and re-written in order to update validation metadata with the new fp chance. However, the loaded metadata lacked compaction metadata, so when rewritten, compaction metadata got truncated. (DB-5005)

• Fixed nodetool not able to setstreamthroughput and setinterdcstreamthroughput (DB-4940)

• Updated the python driver version used by cqlsh from 3.24.0 to 3.25.0 (DB-4978)

18 May 2021

• Apache Solr™ 6.0.1.4.2840
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.5.65
• DSE Java Driver 1.10.0-dse+20210424 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• AOSS returns additional parameter in status endpoint: "connection_hostname". The new parameter is a FQDN of the node hosting AOSS, it may be used for connections (instead of connection_address) if needed. (DSP-21811)

• Fixed an issue with DSE daemon being unable to stop after the default timeout expired. This issue only affected systems that use package install and init.d, such as centos. (DSP-21804)

• Fixed a problem where the Gremlin phrase() predicate may not match Solr results for equivalent search. (DSP-21724)

• A new JVM option is added: dse.search.fc.warmup: AUTO, ALWAYS & NEVER. (DSP-21813)

26 April 2021

• Apache Solr™ 6.0.1.4.2840*
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.5.65*
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed an error in cqlsh encoding unicode in multi-line statements (DB-4855)
• Make cqlsh prefer newer TLS versions. (DB-4966)

• Address a problem where new or rebooted nodes may not be able to gossip with peers. (DSP-21753)

• Upgrade apache commons-compress to address CVE-2019-12402 (DSP-21679)
• Update tomcat version 8.5.61 to 8.5.65 (DSP-21798)

• Fixed a bug where under heavy load solr query worker threads would use 100% CPU due to contention on thread local map (DSP-21746)

9 April 2021

• Apache Solr™ 6.0.1.4.2814
• Apache Spark™ 2.4.0.18
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.5.61*
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.50*

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixes an issue where a login attempt with missing credentials logged a misleading warning message with stack trace instead of an error message about the missing username or password. (DB-4806)

• Fixes a bug when in rare cases a terminated repair session would leak on-heap memory (DB-4833)

• Print a timestamp when nodetool exits due to an error (DB-4826)

• Data export from cqlsh is now less noisy in the logs (DSP-21494)
• Fixes intermittent ERROR: java.util.ConcurrentModificationException at org.apache.cassandra.transport.CBUtil.writeStringList (DSP-21336)
• Fix for DESCRIBE TYPES in cqlsh (DSP-21667)
• Add asynchronous update to KMIP key cache to fix blocking of commit log (DSP-20582)
• Fixes CVE-2020-1945 affecting Apache Ant (DSP-21716)
• Fixes SRCCLR-SID-22742: Insecure Input Validation Vulnerability in the Apache Commons Codec library (DSP-21747)
• Update Tomcat version 8.0.53 to 8.5.61 (fixes CVE-2002-0493 CVE-2009-3548 CVE-2013-2185 CVE-2016-1240 CVE-2016-5018 CVE-2016-5388 CVE-2016-6796 CVE-2016-6797 CVE-2016-8745 CVE-2016-9774 CVE-2016-9775 CVE-2020-8022) (DSP-21394)

• Fixes a severe issue where flushing an empty MemtableIndex causes the index to not be queryable (DB-4934)

• Fixes a bug where FilterCache warmup triggered by node health change can block GossipStage-1 thread for several seconds (DSP-21674)

• Fixes CVE-2014-0114, CVE-2014-0114 (DSP-21668)

11 March 2021

• Apache Solr™ 6.0.1.4.2814
• Apache Spark™ 2.4.0.18*
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Works around a bug in JDK 1.8u282 (JDK-8260018) (DB-4884)

• Fixes a Null Pointer Excpetion in Gossip when upgrading from 5.1 to 6.8.6 (DB-4810)

• During package upgrade yum and apt managers overwrite unedited old jvm.options file. (DB-4705)

• Fixes a problem where FSReadError during streaming could causes DSE to shutdown (DB-4878)

• Port fix of CVE-2020-17516 onto DSE 6.8 (DB-4923)

• Dropped messages metrics calculation doesn't cause assertion errors when dropped messages contain remote batch mutation. (DB-3905)

• SSTablePartitions tool will no longer fail with "histogram overflowed" when its working for the server code (DB-2952)

• Fixes a problem where sstablescrub could not fix a corrupted file (DSP-21672)

• Addressed several Jackson databind vulnerabilities by upgrading jackson-databind to version 2.9.10.8 in DSE 5.1.21, 6.0.15 and 6.7.13 and version 2.10.5.1 in DSP 6.8.10. (DSP-21503) (DSP-21503)
• Fixes a problem where nodetool rebuild could fail intermittently with zerocopy streaming enabled (DSP-21564)

• Update Jetty to 9.4.34.v20201102 and update Spark Versions: DSE 5.1: 2.0.2.38; DSE 6.0: 2.2.3.16; DSE 6.7: 2.2.3.16; DSE 6.8: 2.4.0.17 (DSP-21506)
• SCC by default enables direct join optimization only when size_estimates for both tables are available. (DSP-21628)
• Fix: Spark Master fails to start if keystore (used by web UI) contains more than one certificate (DSP-21703)

• Both graph engines now accept either byte[] or ByteBuffer for blob-typed property values. (DSP-21643)

• Index segments are now merged into a single segment, after the index build. (DSP-19608)

• Fixes a problem where lucene threads were getting interrupted, causing problems with solr cores (DSP-21339)
• Search queries will no longer fail when querying clustering columns of certain types on which the order is reversed (DSP-21363)

• Spark Cassandra Connector supports Storage Attached Indexes (SAI). The connector pushes down predicates defined on columns with SAI indexes. (DSP-21655)
• DSE Spark supports connections to Astra clusters (DSP-21510)

7 January 2021

• Apache Solr™ 6.0.1.4.2814
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Addresses a severe issue where streaming an older file format to a new node could crash sending and receiving nodes. (DB-4846)

• Add support for multiple authorization sources (LDAP + DSE Internal) (DSP-14233)

• Fixed direct join optimization for spark sql. (DSP-21498)

15 December 2020

⚠️ NOTE: Due to a serious bug which affects DSE 6.8.7 and DSE 6.8.8, these releases have been retracted. We recommend against upgrading to these versions at this time. If you have already upgraded to these versions, please EITHER set zerocopy_streaming_enabled=false in the cassandra.yaml and perform a rolling restart AND/OR run upgradesstables on all nodes in your cluster before adding new nodes, running repair, or restoring from backups. This bug has been addressed in DSE 6.8.9. All features and fixes for 6.8.8 and 6.8.7 are present in 6.8.9.

• Apache Solr™ 6.0.1.4.2814*
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• During Backup Service startup, to avoid stalling during service initialization, check cluster readiness in response to endpoint "on alive" events. (DB-4818)

• Fixed an issue in DSE that prevented the Management API to work with DSE versions 6.8.5 - 6.8.7 (DSP-21607)

• The issue: When flushing text column indexes, the internal ordering of terms can be processed out of order causing an "*java.lang.AssertionError: Incremental trie requires sorted keys*" error. When this happens, all flushing of indexes involved in this transaction is aborted and the indexes are marked non-queryable. Recovering from this issue involves either rebuilding the indexes or restarting the nodes. (DSP-21580)
• Fixes a performance regression in SAI for versions 6.8.6 and 6.8.7 regarding *MultiRangeReadCommand*. (DSP-21601)

• A system property dse.solr.fuzzy.max.expansion was added which allows the user to define a custom number of fuzzy query expansions. The maximal possible value is 1024. When unset, the default number of max expansions is 50. (DSP-21605)

• Adjust available framework values for --framework parameter. (DSP-21500)

23 November 2020

⚠️ NOTE: Due to a serious bug which affects DSE 6.8.7 and DSE 6.8.8, these releases have been retracted. We recommend against upgrading to these versions at this time. If you have already upgraded to these versions, please EITHER set zerocopy_streaming_enabled=false in the cassandra.yaml and perform a rolling restart AND/OR run upgradesstables on all nodes in your cluster before adding new nodes, running repair, or restoring from backups. This bug has been addressed in DSE 6.8.9. All features and fixes for 6.8.8 and 6.8.7 are present in 6.8.9.

• Apache Solr™ 6.0.1.4.2794
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixed a bug where the slow query log would fill with queries that do not meet the slow query threshold (DSP-21417)

• Fixed a bug where a single partition read might fail if the following conditions were true:
  1. several sstables had the same partition level deletion info
  2. some of the sstables had wide rows whereas others had not
  3. the sstables in question contained range tombstone markers (DSP-21346)
• Fixed a bug in cassandra.repair.mutation_repair_rows_per_batch setting that caused sending all repair mutations at once (DSP-21429)
• There is a change in SSTable format and/or version. Please refer to the compatibility documentation for more details.

• On the CQLSH CREATE CUSTOM INDEX ... WITH OPTIONS statement, SAI adds support for an ascii option. If set to true, converts alphabetic, numeric, and symbolic characters that are not in the Basic Latin Unicode block (first 127 ASCII characters) to their ASCII equivalent, if one exists. For example, the filter changes à to a. The default is false. (DSP-21409)
• Make the SAI read path synchronous. (DSP-21451)
• Fixed "java.lang.ArithmeticException: integer overflow" printing in system.log when retrieving the SAI index segmentRowID (DSP-21522

• A meaningful error message is logged when two properties with the same name but different types are used in a single core graph. Classic graph was not affected. (DSP-21490)

• Optimized retrieval when memberof_search used the wrong attribute to retrieve groups of the user. (DSP-21537)

• Multi-datacenter backup and restore, new CompositeStore type of backup store. (DB-4489)
• Adds the possibility to restore a backup marked as INCOMPLETE by using the new FORCE RESTORE statement.

• Addressed a bug where a "CommitLogReplayException" is caused by a bad header but correct CRC after restart (DB-3996)
• Fixed a bug where some part of the commit log might not be replayed after injecting a foreign sstable to a node or, on 6.8, after zero-copy streaming of an sstable (DB-4629)

• Fixed an issue where zero copy streaming could cause file descriptor leakage (DB-4594)

• SStableloader now uses native_transport_port_ssl over native_transport_port when passed a config file with the property set (DB-4632)

• Fixed memory leak in Netty resulting in OOM. (DB-4664)
• Fixed a problem in the scheduling and counting of active materialized view updates that could cause too many to be executed concurrently, overwhelming the node. (DB-4782)

12 November 2020

• Apache Solr™ 6.0.1.4.2794
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• A node may be stuck in repair while joining the cluster if broadcast_address is set differently than local_address (DB-4786)

20 October 2020

• Apache Solr™ 6.0.1.4.2794
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Server side backup and restore now supports Microsoft Azure cloud storage as a backup target. (DB-3894)
• snapshot schema.cql files will now contain IF NOT EXISTS clause for CREATE TYPE statements (DB-4685)

• Fixes a problem where races in notifying compaction strategies of added and removed sstables can cause compaction to try to use non-existing sstables and repeatedly fail to make progress. (DB-4711)

• Fixed node restart issue after dropping a PointType column. (DSP-21326)
• Fixed extreme local pauses on all nodes in the cluster on a node restart. (DB-4657)

• Improves performance of estimation of partition counts for subranges. (DB-3679)

• Fsync nodes metadata to prevent FSReadError issues on startup. (DB-4672)

• Fixes LDAP user permissions problem following LDAP server restart. (DSP-21284)

• Fixes LDAP user permissions problem following LDAP server restart. (DSP-21284)

• Escape single-quotes in certain graph-search query predicates. (DSP-21450)

• Fix: Spark Application contacting Nodes in Non Local DC (DSP-19961)

DataStax Enterprise (DSE) 6.8.4 includes all changes from previous DSE versions. See TinkerPop upgrade documentation for all changes.

17 September 2020

• Apache Solr™ 6.0.1.4.2794
• Apache Spark™ 2.4.0.16
• Apache TinkerPop™ 3.4.5-20200107-6cec00d8
• Apache Tomcat® 8.0.53
• DSE Java Driver 1.10.0-dse+20200217 (DSE internal-only version)
• Netty 4.1.25.7.dse
• Spark JobServer 0.8.0.49

NOTE: above-listed DSE Java Driver is an internal-version only. If you're developing applications, please refer to the Java Driver documentation to choose an appropriate version.

• Fixes compaction getting stuck on acquiring references for non-existing sstables. (DB-4290)

• Distributes Netty connections more uniformly across TPC cores (DB-4683)

• Distributes Netty connections more uniformly across TPC cores (DB-4683)

• Adds TTL and TimeWindowCompactionStrategy (TWCS) to system_distributed.repair_history and system_distributed.parent_repair_history tables. (DB-2009)
• DNS Service Discovery is now a part of the DSE/LDAP integration. (DSP-11450)
• New system property to cap the maximum amount of memory used by bloom filters: -Dcassandra.max_bf_memory_mb}. By default, this is unlimited. (DSP-21344)

• DNS Service Discovery is now a part of the DSE/LDAP integration. (DSP-11450)

• DSEFS waits for a schema agreement before starting and issuing the first CQL query. (DSP-20743)

• Storage-Attached Indexing (SAI) adds support for creating multiple SAI indexes on the same collection map column. See SAI collection map examples with keys, values, and entries. (DSP-21306)

DataStax Enterprise (DSE) 6.8.4 includes all changes from previous DSE versions. See TinkerPop upgrade documentation for all changes.

Release notes for previous DSE patch releases can be found here: https://docs.datastax.com/en/dse/6.8/dse-admin/datastax_enterprise/releaseNotes/RNdse.html#RNdse