[jira] [Updated] (TC-199) Configurable consistent hash attributes for cache selection

["Jeff Elsloo (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TC-199?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jeff Elsloo updated TC-199:
---------------------------
    Description: 
Add a feature to Traffic Router and Traffic Ops that allows one to specify attributes to use when consistent hashing for cache selection on a per-delivery service basis. One specific use case is to allow us to consider the query string, as the default is to ignore it. Beyond that, we might want to allow additional strategies, such as a regex that will match parts of the attribute (path, header), or allow URL + header, header only, etc.

Other options:
* short hostname
* domainname
* fully qualified domainname
* Partial path
* Path segments/delimiter (;)
* Headers
* Combinations of the above
* Regular expressions applied to the URL or header(s)

  was:Add a feature to Traffic Router and Traffic Ops that allows one to specify attributes to use when consistent hashing for cache selection on a per-delivery service basis. One specific use case is to allow us to consider the query string, as the default is to ignore it. Beyond that, we might want to allow additional strategies, such as a regex that will match parts of the attribute (path, header), or allow URL + header, header only, etc.


> Configurable consistent hash attributes for cache selection
> -----------------------------------------------------------
>
>                 Key: TC-199
>                 URL: https://issues.apache.org/jira/browse/TC-199
>             Project: Traffic Control
>          Issue Type: New Feature
>          Components: Traffic Ops, Traffic Router
>    Affects Versions: 2.1.0
>            Reporter: Jeff Elsloo
>            Priority: Minor
>              Labels: configuration, routing
>
> Add a feature to Traffic Router and Traffic Ops that allows one to specify attributes to use when consistent hashing for cache selection on a per-delivery service basis. One specific use case is to allow us to consider the query string, as the default is to ignore it. Beyond that, we might want to allow additional strategies, such as a regex that will match parts of the attribute (path, header), or allow URL + header, header only, etc.
> Other options:
> * short hostname
> * domainname
> * fully qualified domainname
> * Partial path
> * Path segments/delimiter (;)
> * Headers
> * Combinations of the above
> * Regular expressions applied to the URL or header(s)



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)