I've read an idea about using a 'ticket' system... each session gets X # of tickets. Tickets regenerate at a fixed rate. Normal users would never run out of tickets. Each query operation would have a fixed cost of tickets. Inserts would cost double selects... You don't have to calculate regeneration - just when an operation is about to be performed, you check the last time a request was made, and the last number of tickets. calculate regeneration and then if they have enough tickets, you do the request. If not, you return a 503 error, or perhaps a friendly message saying "swiper no swiping" (to quote Dora)
On Thursday, May 9, 2013 11:58:43 AM UTC-7, Alex Glaros wrote: > > What techniques can be used in a Web2py site to prevent data mining by > harvester bots? > > In my day job, if the Oracle database slows down, I go to the Unix OS, see > if the same IP address is doing a-lot-faster-than-a-human-could-type > queries, and then block that IP address in the firewall. > > Are there any ideas that that I could use with a Web2py website? > > Thanks, > > Alex Glaros > > -- --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
