Nutch performance

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

Nutch performance

Anton Potekhin
Hello! i would like to know how many pages can nutch index daily and how
many searches can it handle. I understand that it depends on the
hardware because i want to know not exact results ;-) .  For example  i
will use 4 servers.
And i will use the following configuration:
1) server1 i will use for Jobtracker, namenode
2) server2 i will use for the first Tasktracker and the second DateNode
2) server3 i will use for the second Tasktracker and the second DateNode
2) server4 i will use for tomcat for searching

How many pages can i index daily and how many searches can this
configuration handle?

I realize a lot of this depends on the hardware but in general what
would you say. And what can you say what i must change in this
configuration? And what hardware do you recommend to use for each server?