Request limit exceeded when deploying many applications to one machine

[John A Meinel]

I was doing some scale testing, by deploying 15 machines in AWS and then deploying hundreds of units to each machine. While digging through the logs, I see this:
2017-04-25 11:54:15 ERROR juju.provider.ec2 environ.go:864 failed to get instance "i-0b67a6f502cee1309" interfaces: Request limit exceeded. (RequestLimitExceeded) (retrying)
2017-04-25 11:54:15 ERROR juju.provider.ec2 environ.go:864 failed to get instance "i-073e8af1a0d2ba01a" interfaces: Request limit exceeded. (RequestLimitExceeded) (retrying)
2017-04-25 11:54:19 ERROR juju.provider.ec2 environ.go:864 failed to get instance "i-0b67a6f502cee1309" interfaces: Request limit exceeded. (RequestLimitExceeded) (retrying)

Now, there are only 15 machines at play, so I'm not sure what request we are making. Maybe its something like "for each unit find the host machines IP address by querying AWS", or possibly "when targeting a specific machine, ask AWS information about that machine". However, we *should* have all that information already available, as the machine is in use already.

[Canonical Juju QA Bot (juju-qa-bot)]

This bug has not been updated in 5 years, so we're marking it Expired. If you believe this is incorrect, please update the status.