Simba Apache Cassandra ODBC Driver

Transcription

1 Simba Apache Cassandra ODBC Driver with SQL Connector Released These release notes provide details of enhancements, features, and known issues in Simba Apache Cassandra ODBC Driver with SQL Connector 2.2.0, as well as the version history. The following issues have been resolved in Simba Apache Cassandra ODBC Driver with SQL Connector bit driver fails to connect to Cassandra server On heavy load, driver returns "All I/O threads are busy" error Known Issues The following are known issues that you may encounter due to limitations in the data source, the driver, or an application. Insert multiple rows in a single SQL statement and bind DEFAULT value in the first row will be failed. When inserting multiple rows in a single SQL statement and binding the DEFAULT value in the first row, the INSERT statement will be failed. As a workaround, use NULL instead of DEFAULT. Non-key virtual table columns are not nullable. Cassandra does not support null values in collections. The driver reports non-key columns in virtual tables as not nullable. 1

2 Version History Version Released Case Sensitive You can now specify whether to differentiate between capital and lower-case letters in schema, table, and column names. Version History Version Released Support for Cassandra server 2.1 The driver supports read/write for all Cassandra data types except User Defined Types on Cassandra server 2.1. For User Defined Types, the driver can only read the type, and only as a string. Load balancing policies You can now configure the cluster to use either a Round Robin or a DC Aware load balancing policy. Token aware A token-aware option to improve load balancing and latency is now available. Latency aware A latency-awareness algorithm to distribute the load away from slowerperforming nodes is now available. 2

3 Enable Null Value Insertion You can now specify how to handle null values in certain INSERT statements. Connector Failed to passdown multiple count(*)/count(1)/sum(1) aggregation in a single SQL statement Failed to get next page when using filters in an UPDATE statement Truncate a virtual table Version Released Connector When logging is enabled, the driver stops working Version Released Client-server SSL verification implemented You can now configure one-way and two-way verification between the client and the Cassandra server over SSL. 3

4 Version Released Cassandra client API 1.0 GA now being used The driver now uses the 1.0 GA version of the Cassandra client API. Version Released Cassandra client API beta 5 now being used The driver now uses the beta 5 version of the Cassandra client API, instead of beta 3. Support added for ODBC 3.80 The driver now supports ODBC Version Released Connector When you insert DOUBLE data into a table, the inserted data is corrupted. 4

5 Version Released Connector When a prepared query has a WHERE clause that contains a parameter, and the parameter has been bound multiple times, the query returns an error. Version Released Separator for virtual table names now configurable You can now specify the separator to be used in virtual table names by using the "Virtual table name separator" option (the VTTableNameSeparator key). The default separator is _vt_ The name of a virtual table consists of the name of the original table, then the separator, and then the name of the collection from which the virtual table was built. For example: OriginalTable_vt_CollectionName Support added for the "Show Default Schema Only" option in Informatica Developer When using the driver in Informatica Developer, you can now use the "Show Default Schema Only" option in the Connection Explorer. Version Released Data consistency can now be configured through the driver You can now specify a particular replica or the number of replicas that must process a query in order for the query to be considered successful. To configure 5

6 this feature, use the "Tunable Consistency" configuration option (the TunableConsistency key). Large result sets can now be split into pages You can now configure the driver to split large result sets into pages. To configure this feature, use the Enable Paging check box (the EnablePaging key) and the Rows Per Page field (the RowsPerPage key). Version Released User Name and Password authentication added You can now configure the driver to request user login when accessing a Cassandra server. Host connection attribute can now accept multiple servers You can now specify a comma-separated list of IP addresses or hostnames in the Host attribute instead of specifying one server in the Host attribute and then specifying the rest in the SecondaryHosts attribute. The SecondaryHosts attribute is no longer available. The driver will attempt to connect to all the servers concurrently, and then keep the first connection that is successfully established. The driver does not maintain a connection with any of the other servers in the list. Support added for non-ascii Cassandra identifiers You can now use non-ascii Cassandra identifiers. Optimized filter passdown Optimizations for filter passdown functionalities have been implemented. Updated virtual table name delimiter The delimiter for virtual table names has been changed from _vt_ to #vt_ 6

7 Connector The Linux version of the driver does not insert TIMESTAMP data correctly Driver does not detect changes made to the table schema or tables that were created or deleted after the connection was established Collections cannot be displayed when the driver uses CQL_QUERY_MODE This issue has been resolved. You can now retrieve collection columns in CQL_QUERY_MODE. Collection column data is returned as strings. Case-sensitive Cassandra identifiers are not handled correctly Virtual tables do not show foreign keys metadata correctly Version Released Connector IPv6 data is not retrieved correctly Version Released Connector When writing a large amount of data to Cassandra, the driver returns the error message "Batch with conditions cannot span multiple partitions" When writing a large amount of data to Cassandra, the driver returns the error message "Cannot achieve consistency level QUORUM" 7

8 Query fails to execute when the SQL statement does not use any column references In the Windows version of the driver, the length of INET data is not reported correctly, causing equality filters to fail When the connection fails, the driver terminates unexpectedly Version Released Connector The default query mode is not SQL to CQL fallback In some cases, TIMESTAMP values are not correctly written or retrieved The TRUNCATE TABLE query returns an error The CREATE/DROP TABLE query returns an error Catalog names are not case-sensitive INET and UUID data types are not correctly written to Cassandra The driver fails to delete all the data from large virtual and non-virtual tables Version Released Connector When retrieving the primary keys of a table, a memory allocation error occurs 8

9 Version Released SQLForeignKeys supported for virtual and non-virtual tables You can now use SQLForeignKeys in virtual and non-virtual tables. UPDATE and DELETE statements supported for virtual and non-virtual tables You can now use the UPDATE and DELETE statements on virtual and nonvirtual tables. List index columns marked as non-nullable List index columns are now marked as non-nullable. Support added for inserting or updating data at a specific position in a list You can now specify the position in a list in which to insert or update data by writing to the _index column. Connector The driver fails to connect when there is a table with a clustering key in descending order Column order is sometimes inconsistent with the order displayed by cqlsh The driver may terminate unexpectedly when a collection column contains multiple null or empty values 9

10 Version Released Support added for Cassandra 2.0 The driver now supports Cassandra 2.0. Virtual tables feature added Version To support the renormalization of collections, the driver is now able to create virtual tables from list, set, and map columns inside the Cassandra database. Virtual tables allow you to select and insert data as if the data was stored in SQL format. If you insert data into a list at index number -1, the data is appended to the list. If you insert data into a set or map, the data is inserted based on the default ordering in Cassandra. Released Connector The OK button is disabled in the connection dialogue if the DSN text field is empty even if all the required field has been entered. Version Released Writeback Writeback is now enabled. 10

11 Version Released Connector Driver does not use the latest version of Thrift Client. Version Released Version was the initial release of Simba Apache Cassandra ODBC Driver with SQL Connector. 11

12 Contact Us If you have difficulty using the Simba Apache Cassandra ODBC Driver with SQL Connector, please contact our Technical Support staff. We welcome your questions, comments and feature requests. Technical Support is available Monday to Friday from 8 a.m. to 6 p.m. Pacific Time. Important: To help us assist you, prior to contacting Technical Support please prepare a detailed summary of the client and server environment including operating system, product version, patch level and configuration. You can contact Technical Support via: support@simba.com Web site: Telephone: (604) Extension 3 Fax: (604) You can also follow us on 12