[
https://issues.jboss.org/browse/TEIID-3146?page=com.atlassian.jira.plugin...
]
Steven Hawkins edited comment on TEIID-3146 at 1/16/15 10:19 AM:
-----------------------------------------------------------------
I'm not sure we can provide a turn-key billing mechanism. It's best to
concentrate on the underlying parameters and work from there. We have
TEIID-3147 - for bytes transferred to the Teiid client
The various processing time in the query plan and TEIID-2813 which is applicable to
processor threads (although they end up doing some processing work on behalf of other
threads in the buffer manager) as well to get the cpu time rather than the wall clock
time.
We could also look at adding memory batch statistics to the plan - however that can get
messy as typically you'd want to correlate to disk usage which is handled lower down
and by other threads etc.
was (Author: shawkins):
I'm not sure we can provide a turn-key billing mechanism. It's best to
concentrate on the underlying parameters and work from there. We have
TEIID-3146 - for bytes transferred to the Teiid client
The various processing time in the query plan and TEIID-2813 which is applicable to
processor threads (although they end up doing some processing work on behalf of other
threads in the buffer manager) as well to get the cpu time rather than the wall clock
time.
We could also look at adding memory batch statistics to the plan - however that can get
messy as typically you'd want to correlate to disk usage which is handled lower down
and by other threads etc.
Billing mechanism based on the query's complexity
-------------------------------------------------
Key: TEIID-3146
URL:
https://issues.jboss.org/browse/TEIID-3146
Project: Teiid
Issue Type: Feature Request
Reporter: Joao Viragine
Assignee: Steven Hawkins
Priority: Minor
Fix For: Open To Community
It would be nice if we had a billing mechanism based on the query's complexity.
Teeid exposes information about the queries issued, the amount of records returned, the
time elapsed and the query plan.
It could also expose the amount of data (in bytes) returned by each query.
We could combine the above data plus the query complexity to create a score for a billing
calculation.
--
This message was sent by Atlassian JIRA
(v6.3.11#6341)