daos icon indicating copy to clipboard operation
daos copied to clipboard

DAOS-11278 docs: add "Tools for debugging network connectivity issue"

Open wiliamhuang opened this issue 3 years ago • 10 comments

Provide a list of tools for debugging network connectivity issue Provide a list of tools for measuring network latency and bandwidth

Required-githooks: true

Signed-off-by: Lei Huang [email protected]

wiliamhuang avatar Aug 17 '22 13:08 wiliamhuang

Bug-tracker data: Ticket title is 'To add "Tools for debugging network connectivity issue" in docs/admin/troubleshooting.md' Status is 'In Review' Errors are Unknown component https://daosio.atlassian.net/browse/DAOS-11278

github-actions[bot] avatar Aug 17 '22 13:08 github-actions[bot]

A small tip for doc update: it is a good idea to include "Doc-only: true" in the commit to skip unnecessary CI testing.

wangshilong avatar Aug 17 '22 14:08 wangshilong

A small tip for doc update: it is a good idea to include "Doc-only: true" in the commit to skip unnecessary CI testing.

Thanks you! Right. I forgot it in the first commit and added it immediately in the following commit. :)

wiliamhuang avatar Aug 17 '22 14:08 wiliamhuang

@frostedcmos @Michael-Hennecke Could you please review? Thank you very much!

wiliamhuang avatar Sep 01 '22 20:09 wiliamhuang

@frostedcmos @Michael-Hennecke Could you please review? Thank you very much!

Alex is still on Vacation until Oct...Michael might help review.

wangshilong avatar Sep 02 '22 01:09 wangshilong

Looks useful for me, i guess one good thing to have could be some common tools to benchmark network latency and bandwidth except lnet_selftest.

Thanks! ok. I can add example commands for some other related tools.

wiliamhuang avatar Sep 02 '22 04:09 wiliamhuang

@Michael-Hennecke Could you please review it? Thank you!

wiliamhuang avatar Sep 09 '22 12:09 wiliamhuang

What we are still missing is a tool that goes beyond simple point-to-point tests... All the tools mentioned here only check a single p2p connection. That's generally useful, but too cumbersome to systematically test a larger installation...

Thank you! I will add some info about IMB-P2P and intel cluster checker then.

wiliamhuang avatar Sep 14 '22 19:09 wiliamhuang

Added info for the tools to diagnose network issues for a large cluster.

wiliamhuang avatar Oct 07 '22 19:10 wiliamhuang

@Michael-Hennecke Could you please review this PR? Thank you!

wiliamhuang avatar Oct 11 '22 02:10 wiliamhuang

@Michael-Hennecke @mchaarawi Could you please review this PR for docs? Thank you!

wiliamhuang avatar Oct 18 '22 12:10 wiliamhuang

@daos-stack/daos-gatekeeper please expedite the landing of this PR. thank you!

ipoddubn avatar Oct 25 '22 16:10 ipoddubn