[
https://issues.apache.org/jira/browse/TIKA-3582?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17434410#comment-17434410
]
Tim Allison commented on TIKA-3582:
-----------------------------------
I think a better way to fix this is to allow per parse timeout requests to
override the default. We'll have to figure out how to propagate this to
AbstractExternalProcessParser(s). Obv, we won't create any breaking changes,
but we'll deprecate that specific setting for tesseract.
One challenge is that the forked server has to crash and restart to guarantee a
timeout for regular parsers, whereas for tesseract and other external process
parsers, it can stop the forked process reliably.
> Tika does not respect a configuration value passed over a HTTP Header
> ---------------------------------------------------------------------
>
> Key: TIKA-3582
> URL: https://issues.apache.org/jira/browse/TIKA-3582
> Project: Tika
> Issue Type: Bug
> Components: server
> Affects Versions: 2.1.0
> Reporter: dataminer.accolade
> Priority: Major
> Attachments: sampleimage.png
>
>
>
> I think the value of TikaServerConfig.TaskTimeoutMillis should be overridden
> for the current request over *X-Tika-OCRTimeoutSeconds* header. The following
> request takes more than 120 seconds.
> *curl -vvv -X PUT -T sampleimage.png http://localhost:9998/tika --header
> "X-Tika-OCRTimeoutSeconds: 600"*
>
> Tesserect is configured with tessdata_best models
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
