can't see screen of skipadmin in my Local machine.

[h-nakae]

my problem is totally same as below link.

https://stackoverflow.com/questions/62337760/how-to-run-mozilla-hubs-on-my-local-machine

I think I followed the instruction, but the URL which is 'https://hubs.local:4000?skipadmin ' returns white page says 'Missing file index.html. Please try again.' as pasted here.

My current emviroment is like this.

Ubunts 18.04
nodeJS = v12.19.0
(PostgreSQL) 11.9
Elixir 1.10.4 (compiled with Erlang/OTP 22)
ansible 2.9.14
python version = 2.7.17
hubs⇒ cloned from the original repository today

The output of reticulum console when accessed the url is like this..

[debug] QUERY OK db=0.1ms decode=1.7ms queue=7.4ms idle=0.0ms
CREATE SCHEMA IF NOT EXISTS ret0 []
[debug] QUERY OK db=0.1ms queue=1.2ms idle=6.1ms
CREATE SCHEMA IF NOT EXISTS coturn []
[debug] QUERY OK db=5.7ms queue=0.1ms idle=7.3ms
ALTER DATABASE ret_dev SET search_path TO ret0 []
[info] Already up
[debug] QUERY OK db=7.6ms queue=0.4ms idle=28.4ms
INSERT INTO coturn.turn_secret (realm, value, inserted_at, updated_at) values ($1, $2, now(), now()) ["ret", "e35e79c7b17b33758fe8147497f2b99f"]
[debug] QUERY OK db=0.6ms queue=0.5ms idle=17.5ms
DELETE FROM coturn.turn_secret WHERE inserted_at < now() - interval '15 minutes' []
[info] Running RetWeb.Endpoint with cowboy 2.6.3 at 0.0.0.0:4000 (https)
[info] Access RetWeb.Endpoint at https://hubs.local:4000
Interactive Elixir (1.10.4) - press Ctrl+C to exit (type h() ENTER for help)
iex(1)> 04:37:30 - info: compiled 6 files into 2 files, copied 2 in 1.0 sec
[info] TLS :server: In state :certify received CLIENT ALERT: Fatal - Certificate Unknown

[info] GET /
[debug] Processing with RetWeb.PageController
Parameters: [UNFETCHED]
Pipelines: [:secure_headers, :parsed_body, :browser]
[debug] QUERY OK source="app_configs" db=0.9ms queue=0.9ms idle=9281.6ms
SELECT a0."app_config_id", a0."key", a0."value", a0."owned_file_id", a0."inserted_at", a0."updated_at" FROM "ret0"."app_configs" AS a0 WHERE (a0."key" = $1) ["features|default_room_id"]
[debug] QUERY OK source="app_configs" db=0.5ms queue=1.0ms idle=8285.3ms
SELECT a0."app_config_id", a0."key", a0."value", a0."owned_file_id", a0."inserted_at", a0."updated_at" FROM "ret0"."app_configs" AS a0 []
[info] Sent 503 in 10040ms

I am waiting for your replying.
Thanks.

[truedat101]

I can confirm the same error on my system:

iex(7)> [info] GET /
[debug] Processing with RetWeb.PageController
  Parameters: [UNFETCHED]
  Pipelines: [:secure_headers, :parsed_body, :browser]
[debug] QUERY OK source="app_configs" db=1.7ms idle=9229.7ms
SELECT a0."app_config_id", a0."key", a0."value", a0."owned_file_id", a0."inserted_at", a0."updated_at" FROM "ret0"."app_configs" AS a0 WHERE (a0."key" = $1) ["features|default_room_id"]
[debug] QUERY OK source="app_configs" db=0.5ms idle=8231.5ms
SELECT a0."app_config_id", a0."key", a0."value", a0."owned_file_id", a0."inserted_at", a0."updated_at" FROM "ret0"."app_configs" AS a0 []
[info] Sent 503 in 10006ms

[truedat101]

The skipadmin is supposed to avoid having to login using the email links . The problem is that it isn't finding the index.html file . I notice an index.html file in the root level of the project. I am having trouble otherwise understanding the phoenix framework arpproach. I suggest we look through https://github.com/phoenixframework/phoenix and figure it out.

[truedat101]

@robinaguilar713 hoping someone with some experience running this project (or with phoenix) chimes in. I'll continue to dip into documentation. Sometimes people post more helpful set up guides on places like youtube.

[truedat101]

One other detail that may be an issue is the setting of the CORS server. Experiment with changing the config so that you are running a local CORS. The other thing not documented is that https://elixir-lang.org/getting-started/mix-otp/introduction-to-mix.html#environments , noting that MIX_ENV is the variable used to choose the environment, although it seems Elixer defaults to dev mode.

[robinkwilson]

To fix the missing index.html issue (worked for me)-> #425

[truedat101]

@robinaguilar713 heads up. Looks like maybe a fix.

[h-nakae]

@robinkwilson
Thanks, it works fine for me.

[rvanlieshout]

This error is to be found in:

  defp try_chunks_for_page(conn, page, source) do
    res =
      if module_config(:skip_cache) do
        PageOriginWarmer.chunks_for_page(source, page)
      else
        Cachex.get(:page_chunks, {source, page})
      end

    case res do
      {:ok, chunks} = res when chunks != nil -> res
      _ -> {:error, conn |> send_resp(503, "Missing file #{page}. Please try again.")}
    end
  end

In my case I was missing cache entries because of an absence of Ret.PageOriginWarmer in prod.exs. Once I added

config :ret, Ret.PageOriginWarmer,
  hubs_page_origin: "https://hubs.local:8080",
  admin_page_origin: "https://hubs.local:8989",
  spoke_page_origin: "https://hubs.local:9090",
  insecure_ssl: true

config :ret, Ret.HttpUtils, insecure_ssl: true

to my prod.esx and once I made sure there was something running on ports 8080 (https://github.com/mozilla/reticulum#3-start-the-hubs-webpack-dev-server) it was all good. Wait for PageOriginWarmer to update the cache after a phx.server boot. If you access it before that's ready you'll still get the Missing file index.html.

This Phoenix application fetches data from other endpoints. These endpoints need to be available. Note that one should never use insecure_ssl-like options in a production environment!

[truedat101]

@rvanlieshout thanks this is the best description of the root cause that I've seen so far.

[albirrkarim]

after i fix the missing index.html issue. im faces this problem

#496 (comment)

Now I solved this issue by doing this.

https://github.com/albirrkarim/mozilla-hubs-installation-detailed