RE: [PATCH net v2 3/3] selftests: forwarding: Add test for load-balancing between multiple servers

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Ido,

> -----Original Message-----
> From: Ido Schimmel <idosch@xxxxxxxxxx>
> Sent: Sunday, 27 August 2023 19:18
> To: Sriram Yagnaraman <sriram.yagnaraman@xxxxxxxx>
> Cc: netdev@xxxxxxxxxxxxxxx; linux-kselftest@xxxxxxxxxxxxxxx; David S . Miller
> <davem@xxxxxxxxxxxxx>; Eric Dumazet <edumazet@xxxxxxxxxx>; Jakub
> Kicinski <kuba@xxxxxxxxxx>; Paolo Abeni <pabeni@xxxxxxxxxx>; David Ahern
> <dsahern@xxxxxxxxxx>; Ido Schimmel <idosch@xxxxxxxxxx>; Shuah Khan
> <shuah@xxxxxxxxxx>; Petr Machata <petrm@xxxxxxxxxx>
> Subject: Re: [PATCH net v2 3/3] selftests: forwarding: Add test for load-
> balancing between multiple servers
> 
> On Fri, Aug 25, 2023 at 11:08:30AM +0200, Sriram Yagnaraman wrote:
> > Create a topology with a host, and a router. The host (veth0) is in
> > the default namespace, and a network namespace is created for the
> > router, the peer veth1 is moved to the router netns. A dummy interface
> > is added inside the router netns, to simulate a network that has two
> neighbors.
> > An ECMP route to a virtual IP (vip) with the two neighbors as the next
> > hop is added.
> >
> > The test uses perf stat to count the number of fib:fib_table_lookup
> > tracepoint hits for IPv4 and the number of fib6:fib6_table_lookup for
> > IPv6. The measured count is checked to be within 15% for the number of
> > packets received on veth1 in the router.
> >
> > See diagram in the test for more information.
> >
> > Suggested-by: Ido Schimmel <idosch@xxxxxxxxxx>
> > Signed-off-by: Sriram Yagnaraman <sriram.yagnaraman@xxxxxxxx>
> > ---
> >  .../testing/selftests/net/forwarding/Makefile |   1 +
> >  tools/testing/selftests/net/forwarding/lib.sh |   5 +
> >  .../net/forwarding/router_multipath_vip.sh    | 255 ++++++++++++++++++
> 
> I believe this test better fits in fib_tests.sh. Here's a diff that uses the previously
> mentioned method [1]. I verified that it fails without the fixes and passes with
> them:
> 
> Before:
> 
> IPv4 multipath list receive tests
>     TEST: Multipath route hit ratio (.15)                               [FAIL]
>     TEST: Single path route hit ratio (.15)                             [ OK ]
> 
> IPv6 multipath list receive tests
>     TEST: Multipath route hit ratio (.15)                               [FAIL]
>     TEST: Single path route hit ratio (.14)                             [ OK ]
> 
> After:
> 
> IPv4 multipath list receive tests
>     TEST: Multipath route hit ratio (.99)                               [ OK ]
>     TEST: Single path route hit ratio (.15)                             [ OK ]
> 
> IPv6 multipath list receive tests
>     TEST: Multipath route hit ratio (.99)                               [ OK ]
>     TEST: Single path route hit ratio (.14)                             [ OK ]
> 

Thanks a lot, it works for me as well. 
I will post v3 with this version of the test then. 




[Index of Archives]     [Linux Wireless]     [Linux Kernel]     [ATH6KL]     [Linux Bluetooth]     [Linux Netdev]     [Kernel Newbies]     [Share Photos]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Samba]     [Device Mapper]

  Powered by Linux