Exadata server crashes when running in a configuration with mixed RDS fragment size (e.g. mixing UEK2/UEK4)
(Doc ID 2462366.1)
Last updated on FEBRUARY 26, 2019
Applies to:Oracle Exadata Storage Server Software - Version 188.8.131.52.7 to 184.108.40.206.8 [Release 12.2]
Oracle Exadata Storage Server Software - Version 220.127.116.11.0 to 18.104.22.168.0 [Release 12.2]
Information in this document applies to any platform.
In an Exadata configuration that contains servers that use different RDS fragment size (i.e. some servers use 4k and some servers use 16K), a server using 16K fragment size running kernel 4.1.12-94.8.1 or higher may crash when an RDS re-connection to a server using 4K fragment size is attempted. Note the following:
- Oracle Linux UEK4 uses 16K fragment size. Exadata versions 12.2 and higher use UEK4, however, only 18.1.5 through 18.1.9, and 22.214.171.124.7 through 126.96.36.199.8, which use kernel >= 4.1.12-94.8.1, are affected.
- Oracle Linux UEK2 uses 4K fragment size. Exadata versions 12.1 and lower use UEK2.
- Oracle Solaris (SPARC and x86) uses 4K fragment size.
The combination of differing RDS fragment size can occur in the following scenarios:
- Some Exadata servers running Exadata version 12.1 or lower are combined with some servers running version 18.1.5 through 18.1.9, or 188.8.131.52.7 through 184.108.40.206.8.
- Solaris-based database servers (SuperCluster or Exadata) are combined with Exadata storage servers running version 18.1.5 through 18.1.9, or 220.127.116.11.7 through 18.104.22.168.8.
The call stack from the crash of the server using 16K fragment size is similar to the following:
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!
In this Document
|Exadata Database Machine|