[jira] [Updated] (NUTCH-2775) Fetcher to guarantee minimum delay even if robots.txt defines shorter Crawl-delay

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Updated] (NUTCH-2775) Fetcher to guarantee minimum delay even if robots.txt defines shorter Crawl-delay

Chris Mattmann (Jira)

     [ https://issues.apache.org/jira/browse/NUTCH-2775?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sebastian Nagel updated NUTCH-2775:
-----------------------------------
    Fix Version/s: 1.17

> Fetcher to guarantee minimum delay even if robots.txt defines shorter Crawl-delay
> ---------------------------------------------------------------------------------
>
>                 Key: NUTCH-2775
>                 URL: https://issues.apache.org/jira/browse/NUTCH-2775
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher, robots
>            Reporter: Sebastian Nagel
>            Priority: Minor
>             Fix For: 1.17
>
>
> Fetcher uses the amount of seconds defined by "fetcher.server.delay" to delay between successive requests to the same server. Servers can request a longer delay using the Crawl-Delay directive in the robots.txt. This was thought to allow servers to set a longer delay. However, I've recently seen a server requesting "Crawl-Delay: 1". The delay is shorter than the default delay and Nutch may indeed now request one page per second. Later this server responds with "HTTP 429 Too Many Request". Stupid. What about ignoring Crawl-Delay values shorter than the configured default delay or a configurable minimum delay?
> I've already seen the [same issue using a different crawler architecture|https://github.com/commoncrawl/news-crawl/issues/24].



--
This message was sent by Atlassian Jira
(v8.3.4#803005)