|
OK, so how do you increase it? If I've followed this thread right, then
according to what you've stated when i get 4096 simultaneous connections per
real server they will die?
How does one increase to say 8192 per real server? I'm only dealing with
around 3k simultaneous connections per server at the moment, but we are
increasing that number far faster than I'd like. Last thing I need is
systems going down due to insufficient entries.
Also, does this only affect DR based clusters? I'm running a NAT based here.
> The limit is the hash table size of the director. Default value is
> 4096. When the director have 4096 connection alive at the same time, it
> will give you the error.
>
> You can define a larger hash table when you compile the kernel.
>
> > 2) Does anyone have ideas on why the RH6.2 machine doesnt give me the
same
> > error.
> >
> > 3) Not related to the above: How can I calculate the total number of
TCP/UDP
> > connections that my cluster can support in VS-DR, and also the number of
> > connections that my realservers can support.
>
> First it is limited by the hash table size at the director.
>
> Second is the memory and CPU speed of the computer in the cluster.
> If not concern about the memory and CPU, the limit is virtual memory of
> the kernel because each connection will consume some memory. I am not so
> sure about this point.
>
> >
> > Thanks
> > Anush
> >
> >
> > ----------------------------
> > anush@xxxxxxx
> >
> >
> >
> >
>
>
>
|
