Thanks for all your help.
The problem is the memory limit in the startup configuration file.
Specifically, the size of the `stat-segment,` like Pim mentioned, could be an
affecting factor. Besides that, the memory size of the main heap should also be
increased. The resulting configuration needs
Hoi,
It could be that it's the stats-seg that is running out. I found that with
>1M routes in FIB, we need a bit more memory. For our routers, startup.conf
has:
statseg {
size 1G
page-size default-hugepage
per-node-counters off
}
And with this:
pim@nlams0:~$ vppctl show memory stats-segment
Hi,
I think you might find the thread in
https://www.mail-archive.com/vpp-dev@lists.fd.io/msg16302.html useful.
--a
> On 15 Sep 2022, at 17:39, yxm...@case.edu wrote:
> Hello,
>
> I am trying to evaluate the forwarding efficiency of VPP with large FIB/RIB,
> like millions of entries.
>
> H
Hello,
I am trying to evaluate the forwarding efficiency of VPP with large FIB/RIB,
like millions of entries.
However, the VPP crashes when inserting routing entries of more than 310k. The
exact number varies across different routing rulesets but should be around
320k. (One test is 319,739, an
