Trying to start a local instance on Windows

[vinchauhan]

Using Docker Toolbox for windows but ran into the following error 👍

$ docker run --publish 7080:7080 --rm --volume ~/.sourcegraph/config:/etc/s
ourcegraph --volume ~/.sourcegraph/data:/var/opt/sourcegraph --volume /var/
run/docker.sock:/var/run/docker.sock sourcegraph/server:2.13.5
Sourcegraph is initializing the internal database... (may take 15-20 seconds)
Setting up postgres failed:

$ mkdir -p /var/opt/sourcegraph/postgresql

$ chown postgres /var/opt/sourcegraph/postgresql

$ su-exec postgres initdb -D /var/opt/sourcegraph/postgresql --nosync
The files belonging to this database system will be owned by user "postgres".
This user must also own the server process.

The database cluster will be initialized with locale "en_US.utf8".
The default database encoding has accordingly been set to "UTF8".
The default text search configuration will be set to "english".

Data page checksums are disabled.

fixing permissions on existing directory /var/opt/sourcegraph/postgresql ... ok
creating subdirectories ... ok
selecting default max_connections ... 100
selecting default shared_buffers ... 128MB
selecting dynamic shared memory implementation ... posix
creating configuration files ... ok
running bootstrap script ... LOG: could not link file "pg_xlog/xlogtemp.18" to
"pg_xlog/000000010000000000000001": Operation not permitted
FATAL: could not open file "pg_xlog/000000010000000000000001": No such file or
directory
child process exited with exit code 1
initdb: removing contents of data directory "/var/opt/sourcegraph/postgresql"
exit status 1

[sqs]

Thanks for the report. Do the tips at the following links help?

• could not open file "pg_xlog/000000010000000000000001 getsentry/self-hosted#88 (comment)
• https://serverfault.com/questions/699966/fail-to-start-postgresql-when-used-with-vagrant

[azmr]

I had this issue as well. There seem to be 2.5 issues with the "Quickstart" script when using Docker Toolbox for Windows, the first 1.5 cause this particular symptom:

1. /etc and /var are not valid directories; change them to something valid:
   a) for config/data, change to /c/path/to/config_or_data (the docker quickstart terminal treats /c/ as C:\) or $HOME/sourcegraph/config and $HOME/sourcegraph/data.
   Source: https://stackoverflow.com/questions/51523420/postgresql-docker-volume-failing-at-windows

   b) for /var/run/docker.sock:/var/run/docker.sock, change to //var/run/docker.sock:/var/run/docker.sock (just prefix with extra slash), I don't know enough about how docker works to tell you why this would fix things.
   Source: https://stackoverflow.com/questions/36765138/bind-to-docker-socket-on-windows

2. the frontend crashes unless you add -e LOGO=false after docker run. With the tag it still emits a lot of error messages but at the end says:
   frontend | Sourcegraph is ready at: http://127.0.0.1:7080
   (Despite it saying this, when I go to http://localhost:7080 or http://127.0.0.1:7080, I get the page not found error from Firefox: Unable to connect. Firefox can’t establish a connection to the server at localhost:7080. and ERR_CONNECTION_REFUSED from Chrome - would you like me to file a separate issue for this?)

Would it be possible to add a message/option on the Quickstart guide for people on Windows 7 etc who can only use Docker Toolbox for Windows (afaik)?

Cheers,
Andrew

EDIT: for the sake of completeness, the command I ended up using was:

docker run \
  -e LOGO=false \
  --publish 7080:7080 --rm \
  --volume ~/.sourcegraph/config:$HOME/sourcegraph/config \
  --volume ~/.sourcegraph/data:$HOME/sourcegraph/data \
  --volume //var/run/docker.sock:/var/run/docker.sock \
  sourcegraph/server:2.13.5

and the remaining error messages were:

[Andrew@Andrew-PC /e/Program Files/Docker Toolbox ]$ docker run   -e LOGO=false   --publish 7080:7080 --rm   --volume ~/.sourcegraph/config:$HOME/sourcegraph/config   --volume ~/.sourcegraph/data:$HOME/sourcegraph/data   --volume //var/run/docker.sock:/var/run/docker.sock   sourcegraph/server:2.13.5
 Sourcegraph is initializing the internal database... (may take 15-20 seconds)
 16:29:28 zoekt-indexserver | 2018/12/04 16:29:28 Post http://127.0.0.1:3090/.internal/repos/list: dial tcp 127.0.0.1:3090: connect: connection refused
 16:29:33      repo-updater | t=2018-12-04T16:29:33+0000 lvl=warn msg="frontend not available at startup (will periodically try to reconnect)" err="frontend API not reachable: context deadline exceeded (last error: Get http://127.0.0.1:3090/.internal/ping: dial tcp 127.0.0.1:3090: connect: connection refused)"
 16:29:33      repo-updater | t=2018-12-04T16:29:33+0000 lvl=eror msg="failed to run repository clone purge" worker=repo-purge error="Post http://127.0.0.1:3090/.internal/repos/list-enabled: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:33           indexer | t=2018-12-04T16:29:33+0000 lvl=warn msg="frontend not available at startup (will periodically try to reconnect)" err="frontend API not reachable: context deadline exceeded (last error: Get http://127.0.0.1:3090/.internal/ping: dial tcp 127.0.0.1:3090: connect: connection refused)"
 16:29:33           indexer | t=2018-12-04T16:29:33+0000 lvl=eror msg="Could not list unindexed repositories" err="Post: Post http://127.0.0.1:3090/.internal/graphql?ListRepos: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:38           indexer | t=2018-12-04T16:29:38+0000 lvl=eror msg="Could not list unindexed repositories" err="Post: Post http://127.0.0.1:3090/.internal/graphql?ListRepos: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:43           indexer | t=2018-12-04T16:29:43+0000 lvl=eror msg="Could not list unindexed repositories" err="Post: Post http://127.0.0.1:3090/.internal/graphql?ListRepos: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:48      query-runner | t=2018-12-04T16:29:48+0000 lvl=eror msg="executor: error fetching saved queries list (trying again in 5s)" error="Post http://127.0.0.1:3090/.internal/saved-queries/list-all: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:48           indexer | t=2018-12-04T16:29:48+0000 lvl=eror msg="Could not list unindexed repositories" err="Post: Post http://127.0.0.1:3090/.internal/graphql?ListRepos: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:53      query-runner | t=2018-12-04T16:29:53+0000 lvl=eror msg="executor: error fetching saved queries list (trying again in 5s)" error="Post http://127.0.0.1:3090/.internal/saved-queries/list-all: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:53           indexer | t=2018-12-04T16:29:53+0000 lvl=eror msg="Could not list unindexed repositories" err="Post: Post http://127.0.0.1:3090/.internal/graphql?ListRepos: dial tcp 127.0.0.1:3090: connect: connection refused"
 16:29:57          frontend |  Sourcegraph is ready at: http://127.0.0.1:7080

EDIT2: ...using the following docker version:

Client:
 Version:       18.03.0-ce
 API version:   1.37
 Go version:    go1.9.4
 Git commit:    0520e24302
 Built: Fri Mar 23 08:31:36 2018
 OS/Arch:       windows/amd64
 Experimental:  false
 Orchestrator:  swarm

Server: Docker Engine - Community
 Engine:
  Version:      18.09.0
  API version:  1.39 (minimum version 1.12)
  Go version:   go1.10.4
  Git commit:   4d60db4
  Built:        Wed Nov  7 00:52:55 2018
  OS/Arch:      linux/amd64
  Experimental: false

[azmr]

Adding this in a second post to maintain the history.

This may be obvious to long-time docker users, but I hadn't used it before today. Also, apologies if my terminology is not quite correct, I haven't really picked up all the vocabulary of this tech area.
Docker Toolbox seems to use docker-machine, which I think adds another layer of indirection.
AFAICT localhost and 127.0.0.1 don't work because they're consumed/encapsulated by docker-machine, which in turn exposes an IP address of its own to your local computer that you can get with the command:

$docker-machine ip
>192.168.99.100

Source: https://stackoverflow.com/questions/39587279/docker-output-shown-at-http-192-168-99-1008000-instead-of-http-0-0-0-080

Going to http://192.168.99.100:7080 brought up Sourcegraph's 'Create Admin Account' form. Success!

Hope this helps :)
Cheers,
Andrew

[sqs]

Thanks for posting this info. We added the -e LOGO=false to https://docs.sourcegraph.com/admin/faq#docker-toolbox-on-windows.

For the other 2 issues you reported, assigning to @ryan-blunden or @ggilmore to update that doc page (and be aware of code changes to simplify this in the future).