aktuelles_4_1920x800.jpg

New: Informix-Release v15.0!

Version 15.0 of Informix was officially released in November 2024. It has been the first major release for more than 5 years.

The surprise factor was that there had been major physical changes in the internal structures of the engine. These big changes forced a few others to be left out. Consequently not all expected features have been realized. But still a whole lot left over. And by changing all these things IBM layed the foundations to future improvements.

Now let's see a brief list of the most important new features:

  • Java: Engine does not include a JRE environment and Java version 11 is needed

 

  • The internal limits were changed. This means that a lot of limits that we're used to are now much higher. No Informix user will have to worry about limits during his career. Here are some examples:
    • A single table fragment/partition can have 140 trillion pages
    • A chunk may be as large as 8 exabytes
    • Data rows per page depends on pagesize but can reach 29122 in 256KB pages
    • Page size can be 2, 4 ,8, 16, 32, 64, 128 and 256 KBytes (maximum was 32)
    • Data rows per fragment/partition is 9.2 quintillion 

 

  • External smart blobs: Smart blobs can now be stored in external filesystems. This can have a significant impact performing backups. A traditional smart blobspace has to exist to store the smart blobs metadata

 

  • Invisible indexes, is a new functionality that allows you to set an index to "invisible". An invisible index will be supported as a normal index (INSERTs, UPDATEs, DELETEs) but will not be used by queries. So it is possible to test a system behavior as if the index did not exist, and revert easily and immediately if the result is poor. An index can also be created in "invisible mode" and turned visible later.

  • Obtain a query plan of a running query. This is important if you have a performance issue to solve. Why is this an issue:
    • It's difficault capturing a query and it's parameters (if the query is prepared), run it in a tool and obtain the query plan. Specially for short lived queries
    • A prepared query may have a different query plan than the one we get in a tool with a query written with values
    • A long running query may be using a different query plan than the one we get, because some of the conditions may be different (update statistics may have been run, parameters or context may have changed)

 

  • Improved audit records analysis. This means the Session ID (SID) can appear in the audit logs. Before in many cases it was not possible to correlate the actions shown in the log.

 

  • Informix HQ 3.0:
    • Automatic account lockout after 5 invalid access attempts
    • Support for "LARGE TABLES". LARGE TABLEs can be created and seen if a table is categorized as SMALL or LARGE table.
    • Droping constraints directly from the user interface
    • Multiple servers refresh simultaneously

 

  • CDC API support on secondary servers. Being able to take advantage of secondary servers will be great news for Informix customers who use CDC to send data to other systems.

 

Find more information here: 

https://www.ibm.com/docs/en/informix-servers/15.0.0?topic=overview-whats-new-in-informix#concept_v15.0.0.0__cascade_repl_15.0.0.0