Suggestions for plugin release

[cbruno10]

Hey @Bonemind and @graza-io , thanks for creating this plugin! We've reviewed it and have some suggestions before publishing below. If you have any questions, please let me know, thanks!

index.md:

• brand_color should be #696566
• For icon_url, you can use images/plugins/theapsgroup/vsphere.svg
• For og_image, you can use images/plugins/theapsgroup/vsphere-social-graphic.png

go.mod

• We've recently released Steampipe SDK v1.6.1. I don't believe this plugin needs any of the features in v1.6.0 and onward, but we'd recommend upgrading as there are some bug fixes in between v1.5.0 and v1.6.1.

vsphere/utils.go

• There are two places where we just print out connection errors with debug like messages (https://github.com/theapsgroup/steampipe-plugin-vsphere/blob/master/vsphere/utils.go#L21, https://github.com/theapsgroup/steampipe-plugin-vsphere/blob/master/vsphere/utils.go#L31) and don't return the error. Were the messages leftover from testing? And is the connect function setup to handle returning errors OK?

README.md

• Not a requirement in the file, but would it be easier if you included a Makefile, similar to how we've done in other plugins (https://github.com/turbot/steampipe-plugin-jira/blob/main/Makefile), and then ask users to run make instead of go build .../mv ... in the alternate instructions?

Table and column descriptions

• Some of the descriptions seem inconsistent or contain typos:

Different casing of datastore, also missing periods at the end of each sentence

  > .inspect vsphere.vsphere_datastore
  +-------------+---------+--------------------------------------------------------------------------+
  | column      | type    | description                                                              |
  +-------------+---------+--------------------------------------------------------------------------+
  | accessible  | boolean | Whether this datastore is accessible                                     |
  | capacity    | bigint  | Capacity in bytes                                                        |
  | free        | bigint  | Free space left in bytes                                                 |
  | name        | text    | The name of the datastore                                                |
  | type        | text    | The type of Datastore                                                    |
  | uncommitted | bigint  | How much storage on this datastore has been allocated to guests in bytes |
  +-------------+---------+--------------------------------------------------------------------------+

Typo in the word Curent:

  > .inspect vsphere.vsphere_host
  +--------------+--------+----------------------------------+
  | column       | type   | description                      |
  +--------------+--------+----------------------------------+
  | cpu          | text   | CPU model                        |
  | cpu_cores    | bigint | CPU core count                   |
  | cpu_mhz      | bigint | CPU clock rate in mhz            |
  | cpu_threads  | bigint | CPU thread count                 |
  | cpu_usage    | bigint | Current cpu usage in mhz         |
  | manufacturer | text   | The manufacturer of the hardware |
  | memory       | bigint | Memory in bytes                  |
  | memory_usage | bigint | Curent memory usage in MB        |
  | model        | text   | The model of the hardware        |
  | name         | text   | The host's name                  |
  | num_hbas     | bigint | The number of HBAs connected     |
  | num_nics     | bigint | The number of NICs connected     |
  | status       | text   | The status of the host           |
  | uptime       | bigint | Uptime                           |
  +--------------+--------+----------------------------------+

Are these descriptions pulled from the vSphere SDK or API, or somewhere else? In general, we prefer to use any descriptions from those sources so we can provide users with accurate, easy to find information. Our (short) list of best practices around table and column descriptions can be found at Table and Column Descriptions.

[cbruno10]

Also, while looking through the table code, the list functions didn't use any paging. Does the SDK support pagination today for some or all of the resource list calls?

[graza-io]

@cbruno10 thanks we'll get on it :)

[Bonemind]

I've merged a pr that should address your points. SDK doesn't seem to support pagination

[cbruno10]

Thanks @graza-io and @Bonemind !

I believe the plugin is close to ready for publishing, but I had a few suggestions/questions on the config options and docs:

• For the config options, any thoughts on using the Terraform config names from https://registry.terraform.io/providers/hashicorp/vsphere/latest/docs#argument-reference, e.g., user, vsphere_sever,allow_unverified_ssl? These seem to be based off of the VMWare standard environment names, e.g.,VSPHERE_USER`, so they may be familiar to current users.
• How does the plugin behave if the configuration file isn't set to allow insecure connections? Do the users need to configure anything else client side to allow the connections to succeed?
• What's the default behaviour if insecure isn't set in the config file today, does it attempt to connect without allowing insecure connections?
• Can you please add 2-3 more examples in each table's doc? We consider these documents key for plugins, as they're very helpful for users looking to learn how to use the plugin/tables on the Hub site.

[Bonemind]

Thanks for the feedback. I've renamed the config variables and allowed them to also be set using environment variables. Insecure connection (now allow_unverified_ssl) is just for the validation of the server's ssl certificate, So no additional configuration for users. I've also defaulted insecure connections to false.

[cbruno10]

Thanks @Bonemind ! I'm closing this issue as v0.0.1 is now released 😃