[BioCatalogue-developers] about blocking an ip

Eric Nzuobontane ericnzuo at ebi.ac.uk
Mon Mar 1 05:51:36 EST 2010

Here is what the systems people told when I asked  the ip below could be 


First of all I would not block this IP because it belongs to an ISP, so
another user could be allocated it.

But why would you want to block it anyway?  It was by far your best 
customer on 26/2/2010, and behaved very well, only using 2 connections 
at a time and making legitimate requests.  We only block access where a 
client program has gone beserk, or is disobeying clearly stated rules. 
I do not see anything banning programmed download of the biocatalogue.

On that day it did make 10K hits, but googlebot made nearly 16K.  Some 
of the google requests download huge amounts of data, so I have now 
blocked dynamic queries in the robots.txt.  They did not appear in the
google index anyway.  I assume you are having problems so I have blocked 
spamming robots that give us no referrals in your robots.txt.

The only problem I see with requests from is that there 
were some timeouts, so I have increased the timeout at the loadbalancer.

I would strongly suggest that you turn off debugging on your production
server.  Your error and production logs are enormous, and writing them 
will slow down your service.


We really should have investigated the IP a bit more. Anyway, the advise 
is that there is no need to block this IP for reasons stated above.

Eric Nzuobontane
European Bioinformatics Institute
Wellcome Trust Genome Campus
Hinxton, Cambridge
CB10 1SD
United Kingdom

Tel   : +44 1223 492654
email : ericnzuo at ebi.ac.uk

More information about the BioCatalogue-developers mailing list