Thanks Dave, in our use case the number of reads are definitely going to be
much higher than the number of writes. That said, we also expect number of
writes to be about 10K - 20K writes/second. I am trying to figure out if any
one has optimized the Bihash table for this type of use case as
Hi,
We have a use case that requires us to implement a hash table with multi-thread
support. In our use case all VPP workers can add/remove/access (i.e. write and
read) entries to/from this hash table. I looked at the VPP Bihash
implementation that has the property of not requiring reader
Hi,
I am using the API vec_sort_with_function to sort a vector and I am noticing
that it crashes. This API uses qsort function to sort the vector. Are there any
knows issues with the implementation of qsort? I did see a couple of messages
in this list discussing potential buffer overflow
On Wed, Sep 11, 2019 at 10:20 AM, Burt Silverman wrote:
>
> Perhaps Krishna could use a total of 2 VLANs rather than 2 VLANs per line
> card, and also think in terms of point to point rather than subnets. I
> have not thought through whether or how that helps hide the traffic
> splitting/load
Our product has multiple line cards and each line card has multiple interfaces.
We run an instance of VPP on each line card. All of these interfaces are
connected to a L2 switching network. A PE router is also connected to this L2
network for connectivity to internet.
In order to explain this
