Boardfarm ping6 test from ci40 to a clicker #21
Conversation
tests/ping6_clicker.py
Outdated
| class BoardPing6ClickerDev(rootfs_boot.RootFSBootTest): | ||
| '''Board lowpan interface can ping6 the clicker.''' | ||
| def runTest(self): | ||
| # Make Lan-device ping Wan-Device |
There was a problem hiding this comment.
What is the purpose of this comment being here?
tests/ping6_clicker.py
Outdated
| '''Board lowpan interface can ping6 the clicker.''' | ||
| def runTest(self): | ||
| # Make Lan-device ping Wan-Device | ||
| board.sendline('\nping6 -I lowpan0 -c 20 fe80:0000:0000:0000:0204:a30b:0000:c184') |
There was a problem hiding this comment.
I think somewhere in the test doc we should mention the distance between the DUT and clicker device, which should be maintain constant throughout.
There was a problem hiding this comment.
We have seen different issues depending on the size of the packets so it would be useful for this test to ping 20 packets of small, medium and large sizes. @mtusnio any suggestions on good sizes, I'm thinking along the lines of the fragmentation limit?
Also should the IP address to ping should come from an env var on the boardfarm slave?
tests/ping6_clicker.py
Outdated
| '''Board lowpan interface can ping6 the clicker.''' | ||
| def runTest(self): | ||
| # Make Lan-device ping Wan-Device | ||
| board.sendline('\nping6 -I lowpan0 -c 20 fe80:0000:0000:0000:0204:a30b:0000:c184') |
There was a problem hiding this comment.
We have seen different issues depending on the size of the packets so it would be useful for this test to ping 20 packets of small, medium and large sizes. @mtusnio any suggestions on good sizes, I'm thinking along the lines of the fragmentation limit?
Also should the IP address to ping should come from an env var on the boardfarm slave?
tests/ping6_clicker.py
Outdated
| board.expect(' ([0-2]+)% packet loss') | ||
| n = int(board.match.group(1)) | ||
| board.expect(prompt) | ||
| assert n > 0 |
There was a problem hiding this comment.
Un-shielded production test JIGs in the lab show in busy environments with hundreds of clickers about a non-zero packet loss isn't always possible. Perhaps we could bump this to 10% to avoid any false negatives?
There was a problem hiding this comment.
Around 80-90 bytes should be good, you can try capturing the traffic and see the overhead of ping header and such to get the right amount, but with UDP around 84 bytes was the fragmentation limit. A lot of it depends on the header size and such though, since if I remember correctly fragmentation cuts off at about 128 bytes.
tjobanpu
force-pushed
the
3_waffle_ci40_clicker_ping
branch
3 times, most recently
from
e662fa5 to
840beaf
Compare
This test will ping6 20 packets from the ci40 board to a clicker using 6lowpan wireless interface Signed-off-by: Tushar Jobanputra <[email protected]>
tjobanpu
force-pushed
the
3_waffle_ci40_clicker_ping
branch
from
840beaf to
e651014
Compare
This test will ping6 20 packets from the ci40 board to a clicker
using 6lowpan wireless interface
connects #3
Signed-off-by: Tushar Jobanputra [email protected]