[
https://issues.apache.org/jira/browse/TS-3178?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14201600#comment-14201600
]
Leif Hedstrom commented on TS-3178:
-----------------------------------
Since this is going to google, they are presumably cache misses? I'd probably
suggest that we do not benchmark against someones production services ;).
Another obvious issue with the two tests is that they seem to be different :).
The first is an average of 300 bytes, and the second an average of 1180 bytes.
Why is that? Also, why is this so slow in general? 1s or more average response
time, where is that coming from? Network? Google throttlign you?
Finally, why are we testing redirects? All tests says
Non-2xx responses: 10000
I asume goggle is giving you a redirect response?
> ProxyAllocator improvements
> ---------------------------
>
> Key: TS-3178
> URL: https://issues.apache.org/jira/browse/TS-3178
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Brian Geffon
> Assignee: Cynthia Gu
> Attachments: patch.diff
>
>
> Currently when a ProxyAllocator (Thread Local) has more than a
> configurable number of elements it will return them one-by-one to a
> ClassAllocator (Global Freelist). Returning every item in this fashion is
> inefficient as we'll likely need more items in the future. Therefore we
> should maintain a low watermark (a minimum number) of items that should be in
> a ProxyAllocator at any one time. When the number of elements reaches high
> watermark, the free up is triggers to keep it below the low
> watermark. Additionally, the free should be a block free instead of
> one-by-one as we can reduce several hundred compare-and-swap operations to a
> single CAS
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
