Thanks for the pointer. It does perfectly the job!
-----Original Message----- From: Dennis Kubes [mailto:[EMAIL PROTECTED] Sent: Tuesday, August 29, 2006 7:59 AM To: [email protected] Subject: Re: Hadoop job question Although it is kinda hacking the system you may be able to do it in the map method by writing a custom MapRunner and having an object that lives in the MapRunner but that you set into each mapper instance. Dennis HUYLEBROECK Jeremy RD-ILAB-SSF wrote: > I currently have a MR task that reads a SequenceFile via the map > method to output some data. > > My goal is to output some data to MySQL but I'd like to read several > records before doing the INSERT. > > But I can't figure out how to get several records... > They all have different keys so the reduce task only gets one at a time. > > > Thanks for any help! > >
