[
https://issues.jboss.org/browse/TEIID-4489?page=com.atlassian.jira.plugin...
]
Kylin Soong commented on TEIID-4489:
------------------------------------
[~van.halbert], If I understand correctly, the following 2 can be used as rules
* if sources are 5-10, federated query rows are in millions, then recommend small( 16
cores, with at least 32 GB RAM. 2 DV instance)
* if sources are 10-20, federated query rows are in multi-millions then recommend large(32
cores with 64 GB RAM, 4 DV instances)
* else base on formula, calculate the results
Note that, current RAM size and core size base on formula, if default recommend ( 16
cores, with at least 32 GB RAM. 2 DV instance), it may not accordant with formula.
Adjust defaults and use of 1 to 10 data sources to recommend the
SMALL size
---------------------------------------------------------------------------
Key: TEIID-4489
URL:
https://issues.jboss.org/browse/TEIID-4489
Project: Teiid
Issue Type: Enhancement
Components: Sizing Application
Affects Versions: 9.x
Reporter: Van Halbert
Assignee: Kylin Soong
Based on recommendations:
small : 16 cores, with at least 32 GB RAM. 2 DV instances. (small means may be 5 ~10
sources, rows into millions, YMMV)
medium : 32 cores with 64 GB RAM, 4 DV instances. (medium means may be 10 ~ 20 sources,
rows into multi-millions, YMMV)
Large: custom
The defaults in the sizing applications, along with selecting 1 to 10 data sources,
should result in the SMALL recommendation. It should be clear to indicate the number of
instances and the size recommendation per instance.
--
This message was sent by Atlassian JIRA
(v6.4.11#64026)