Mercurial > hg > cc > cirrus_home
changeset 71:17eb428525cb
impose some limits
| author | Henry S. Thompson <ht@inf.ed.ac.uk> |
|---|---|
| date | Tue, 28 Apr 2020 19:02:14 +0100 |
| parents | efa138bcb171 |
| children | f1bf3effa893 |
| files | src/nutch-cc/conf/nutch-site.xml |
| diffstat | 1 files changed, 2 insertions(+), 2 deletions(-) [+] |
line wrap: on
line diff
--- a/src/nutch-cc/conf/nutch-site.xml Tue Apr 28 19:01:41 2020 +0100 +++ b/src/nutch-cc/conf/nutch-site.xml Tue Apr 28 19:02:14 2020 +0100 @@ -5,14 +5,14 @@ <configuration> <!-- from email from Sebastian 2020-04-16, ccrawl:57[01] --> - <property><name>http.content.limit</name><value>-1</value></property> + <property><name>http.content.limit</name><value>500000000</value></property> <property><name>http.store.responsetime</name><value>true</value></property> <property><name>store.ip.address</name><value>true</value></property> <property><name>store.http.request</name><value>true</value></property> <property><name>store.http.headers</name><value>true</value></property> <property><name>http.accept.language</name><value>en-US,en;q=0.5</value></property> <property><name>http.accept.charset</name><value> </value></property> - <property><name>http.time.limit</name><value>300</value></property> + <property><name>http.time.limit</name><value>600</value></property> <property><name>http.timeout</name><value>45000</value></property> <property><name>http.redirect.max</name><value>3</value></property> <property><name>http.redirect.max.skip</name><value>false</value></property>