k6 load issue with metalb load-balancer with high tps from external driver

[mradulamd]

Brief summary

Hi ,

k6 load issue with metalb load-balancer with high tps, if we execute with aws load balancer in external driver then it is 100% success.

how to match 100% with metalb load balancer . above 2000 tps it is not supported .

k6 version

k6/0.30.0

OS

linux [ kubernetes ]

Docker version and image (if applicable)

No response

Steps to reproduce the problem

Hi ,

when hit 2000+ tps then packet drop start . we have tried 5000 tps , and 3000 tps is dropping

Expected behaviour

when hit 5000 tps then no backet should drop

Actual behaviour

when hit 2000+ tps then packet drop start . we have tried 5000 tps , and 3000 tps is dropping

[joanlopez]

Hi @mradulamd,

I'm sorry, but from your considerably vague description I cannot fully understand what's going on, plus I'm not familiar at all with MetaLB, which we don't specifically offer support for.

Could you please explain more in detail what do you mean with "packet drop", and why do you think it's k6's fault?
How do you observe the packet drop? What's actually dropping the packets?
Because, from my standpoint, the problem is likely with MetaLB that is unable to handle such high throughput.

That said, it also worths mentioning that you're using a very old version of k6 (v0.30.0 is more than 4 years old!), so the general recommendation here would be to update to the latest version (or close to), because even if this is a problem in k6, which is unlikely, we definitely won't cover that support in such an old version anyway.