ctron / hot

Command line test tool for Eclipse Hono

HoT – Hono Test

This is a simple command line tool for testing Eclipse Hono™.

Authentication

Username and password

You can set the username and password for all operations using the --username and --password parameters.

When publishing data the username is normally a combination of <auth-id>@<tenant>. As you are already providing the tenant, you can use the --auth-id parameter instead, which will internally generate the correct user name, by adding the tenant suffix.

--auth-id,-a

The username to use for authenticating with the backend.

--username,-u

The full username to use for authenticating with the backend.

--password,-p

The password to use for authenticating with the backend.

Assuming you have a tenant foo and an authentication id of auth1, then you can use either:

--username auth1@foo

Or:

--auth-id auth1

X.509 client certificates

It is possible to use X.509 client certificates, instead of username/password authentication. For this you can use the parameters:

--client-key

The path to an X.509 PKCS#8 encoded private key

--client-cert

The path to an file containing a PEM encoded client certificate chain

Start a test consumer

Fill in your connection information, and then execute the following command:

hot consume amqps://my.server:443 tenant

Or:

hot consume amqps://username@password:my.server:443 tenant

You can use the following flags:

--disable-tls

Disable TLS negotiation on the AMQP connection

--insecure

Set to true to enable Insecure TLS connection

--cert

Path to the certificate bundle in PEM format (overrides system CA certs)

Telemetry & event

Running consume will consume both telemetry and event messages and output them on the console.

Command & control

You can enable command and control handling, by using the switch -c, --command. You can pass an optional value to the switch, which is the command name, TEST is being used by default.

When this feature is enabled and a message containing the ttd property is received, it will try to get the next command, and forward it to the device.

The source of the command can be specified with the -r,--reader argument. The following readers are available:

• ondemand – When a command is required, it will show a prompt on the console, which can read the command payload. When the command requests times out, the prompt will get canceled.
• prefill – A prompt on the console allows putting in the command payload for the next command request.
• static:<payload> – Everything after the prefix static: will be used as the command payload. No interactive prompt is being presented.

Publish an HTTP message

Fill in your connection information, and then execute the following command:

hot publish http telemety https://my.server tenant device payload --username auth --password password

The following additional flags are supported:

--qos

Set the "Quality of Service". Defaults to 0.

--ttd

Set the "time till disconnect", the amount of seconds the HTTP call will wait for a command to the device

Publish an MQTT message

Fill in your connection information, and then execute the following command:

hot publish mqtt telemety ssl://my.server tenant device payload

The following additional flags are supported:

--qos

Set the "Quality of Service". Defaults to 0.

--ttd

Set the "time till disconnect", the amount of seconds the MQTT call will wait for a command to the device

Building

Building requires Go 1.13.x. You can build the binary by executing:

GO111MODULE=on go build -o hot ./cmd

hot's Issues

Align TLS options with URLs

Currently we have the problem that you can specify a TLS based endpoint (e.g. amqps vs amqp), which would also enable TLS internally. However, if you want to specify a custom TLS certificate, you need to add -T and -P. Only adding -P doesn't work.

I think when using a TLS based endpoint (same for mqtt and http), then -P or --insecure should be sufficient for configuring TLS.