Improve error message on proxy.checkHealth timeout

[hkraal]

The same goes for the remote check of a proxy.

Context

• XO origin: XO Appliance
• Versions:
  • Node: v12.18.0
  • xo-web: 5.60.0
  • xo-server: 5.60.0

Expected behavior

If the proxy is not reachable due network configurations I would expect a clear message which states that.

Current behavior

When the proxy is unreachable (in this case due an intermediary firewall) and a packet is dropped the following error is shown in settings/logs with a popup that "an unknown error occurred".

proxy.checkHealth
{
  "id": "11ee79c5-a766-43c3-865b-ad01d1ad0b22"
}
{
  "code": "ECONNRESET",
  "url": "https://192.168.0.2/api/v1",
  "message": "socket hang up",
  "name": "Error",
  "stack": "Error: socket hang up
    at connResetException (internal/errors.js:609:14)
    at TLSSocket.socketCloseListener (_http_client.js:400:25)
    at TLSSocket.apply (events.js:327:22)
    at TLSSocket.patchedEmit [as emit] (/usr/local/lib/node_modules/xo-server/node_modules/@xen-orchestra/log/src/configure.js:94:17)
    at net.js:674:12
    at TCP.done (_tls_wrap.js:574:7)"
}

[badrAZ]

Hi,

We cannot be sure that this error is due to the network configuration, it can also be triggered by a lost of connection.

What we are sure when this error is triggered is that the XOA cannot connect to the proxy but we can't know precisely the cause.

[julien-f]

@badrAZ Maybe we could check the error code and have special cases for the most common ones like ECONNRESET, ECONNREFUSED, etc.

[badrAZ]

@julien-f By adding for example a property(cause) which gives more information about these errors?

Regarding these common erros only (ECONNREFUSED, ECONNRESET, ENOTFOUND, ETIMEDOUT) are pertinent in case of a proxy health check.

[julien-f]

No, handle this on UI's side, test code and choose the appropriate message.

Don't hesitate to group errors together if the same message and the same user solution can apply.