You’ll probably need to provide some more information like hardware specs
(cpus/cores, memory, disks), number of nodes in the cluster, JVM heap size,
configuration of repositories, specific processors that are bottle necks,
etc.

On Tue, Oct 23, 2018 at 10:02 PM Phil H <[email protected]> wrote:

> Yep - although it now has less throughput with 8 threads than it
> previously had with one.
>
> I have seen this happen when you have a lot of errors going into the logs,
> but that doesn’t appear to be the case this time around
>
> Run Schedule is set to 0 sec
>
> > On 24 Oct 2018, at 11:38 am, Milan Das <[email protected]> wrote:
> >
> > Did you increase the no. of Concurrent Tasks and Run Duration on slow
> processors ?
> >
> > Thanks,
> > Milan Das
> >
> > On 10/23/18, 8:26 PM, "Phil H" <[email protected]> wrote:
> >
> >    Hi team,
> >
> >    My NiFi is struggling to process data (lots of back pressure in
> various
> >    queues) but it is only using a very small amount (2-5%) of CPU
> according to
> >    top.
> >
> >    Any ideas?
> >
> >
> >
>
> --
Sent from Gmail Mobile

Reply via email to