Re: Question about exceeding high write bytes watermark

[Michael Penick]

Sorry for the delay. I've been out for two week.

It can be set as high as sustainable by your client machines and/or Cassandra cluster can handle. Or it can be set arbitrarily high to disable backpressure entirely.

Mike

On Fri, Jun 5, 2015 at 5:08 PM, Robin Mahony <rob...@netapp.com> wrote:

Also, how high can this value be set safely?

On Friday, June 5, 2015 at 4:46:36 PM UTC-7, Robin Mahony wrote:
> Hi,
>
> So I am exceeding the write bytes high watermark while using the async interface. I have a question about how this works. When I exceed it, I get that it disables the writing to a connection. What exactly does this mean? Should any attempts to perform an 'INSERT' fail with an error? Or does it queue up the data, but just not write it?
>
> Also, once the queue has fallen below the low bytes water mark I assume it should then call the async callback supplied?
>
> It seems like I am exceeding this watermark, but then it is never calling the callback so I wait forever.
>
> Also, I assume it tries to write bytes in the order they are queued? Ie. The first INSERT done should be the first one written once the connection is open again?
>
> Just looking for some clarification here as either I am doing something wrong or there is a bug with exceeding the watermark and async operations.
>
> Cheers,
>
> Robin

To unsubscribe from this group and stop receiving emails from it, send an email to cpp-driver-us...@lists.datastax.com.