Lack of sub-command to query the current disk latency injection status

[innerr]

Background

The kchaos client help info:

$> kchaos -h
kchaos is the client to communicate with chaos driver

Usage:
   [flags]
   [command]

Available Commands:
  completion  generate the autocompletion script for the specified shell
  help        Help about any command
  inject
  recover
  version

Flags:
  -h, --help   help for this command

Use " [command] --help" for more information about a command.

As we can see, the common usage is:

• kchaos inject ..., if it succeeded, it will give us an injected-id as the handle for later recovering
• do something
• kchaos recover --id <injected-id>, it will remove the injection

Issues

We lack of sub-command to query the injection status, hence could lead to problems:

• The injected-id is a monotonic increasing value, after many times of kchaos calls, it could be any value which we have no way to retrive for recovering
• We could inject latency to one target(aka, disk) many times, but we have no way to retrive the info of how many times and what latency it is for each time

Workaround

To recover an injected disk (without previous injecting info):

• do an meaningless kchaos inject ... to get the current and also the largest injected-id
• for n in 0~9, do: kchaos recover --id <injected-id> - n, then PRAY for this operating could remove ALL the injections