I'd like to draw your attention to a new section in the manual on the
"MySQL Cluster Limitations in 4.1" that might be of interest to all:
Please bookmark, read, think and provide feedback.
A couple of months back I posted a question about "Production usage of MySQL Cluster?".
I got some good feedback from some users that I'd like to share as well and comment on.
The following issues/items were raised (all relevant ones of them are also mentioned also
manual section above):
* Case insensitivity in LIKE compares
Supported as of 4.1.6 (introduced collation)
* NULLs in indices
Supported from 4.1.6
* Add and remove nodes without restarting the cluster
This is not supported. Most other configuration parameters can be upgraded on-line. It
will not be supported in 5.0.
* Disk-based NDB tables
We're working on it, althogh we will no have it for 5.0.
* 100% compliance with the set of MySQL-style SQL commands
This is the plan and we're working on it, but not quite there. 4.1 will not be fully
(see list in manual section above).
* Char Sets (UTF8)
Support for some from 4.1.6
* Referential Integrity
No support. Possibly in 5.0.
* Fully searchable indexes (WHERE NAME LIKE 'c%')
Supported, please provide feedback on constructs that you see fuction in a wrong way.
The construct above results in an index search as it should.
'%c' will however result in a full table scan.
* Performance when trying to run complicated filtering/joining queries
We've made quite a few improvements during the fall. See MySQL Cluster news section in
We've scheduled some major improvements for 5.0.
Please take a moment to think about what your're missing, having issues with, regarding
I will shortly follow up with a mail with some questions, regarding this, to the list.
Thank you all,