This post is about deploying an instance of Seafile using Docker on a non-standard port (i.e. not 80 and 443). The reason to deploy the app on a non-standard port is usually because there is already another webserver running on that port (in my case Apache 2). In such scenarios, it is necessary to set up a reverse proxy so the Seafile port can be exposed by Apache on ports 80/443. Under typical circumstances, deploying Seafile on a standard port would mean that the app would host itself without the need for Apache 2 at all. This post also assumes Ubuntu 16.04 as the operating system, but in theory should apply to other Linux OS’s.

The process can be split into multiple parts:

1. Setup Docker
2. Deploy Seafile
3. Configure Apache 2
4. Setup systemd
5. Setup e-mail (optional)
6. Backing up the image
7. Troubleshooting

Setup Docker

This guide won’t go into depth on how to setup Docker, but rather the setup guide and more information about what Docker is can be found here: https://docs.docker.com/get-started/

Deploy Seafile

The key page in the Seafile documentation for deploying Seafile within Docker can be found here: https://github.com/haiwen/seafile-docker. The command I used is as follows:

docker run -d --name seafile \
  -e SEAFILE_SERVER_HOSTNAME=seafile.example.com \
  -e SEAFILE_ADMIN_EMAIL=me@example.com \
  -e SEAFILE_ADMIN_PASSWORD=a_very_secret_password \
  -v /opt/seafile-data:/shared \
  -p 8000:80 \
  seafileltd/seafile:latest

I recommend setting up Seafile with an actual secret password because it’s not easy to change once it’s set up.

Note that the port argument shows 8000:80. It means that Docker will use the host port 8000 and the container will use port 80 – meaning that using Docker will interpret requests from my host OS on port 8000 as requests on port 80 in the container. 8000 was used on my setup because port 80 was already used.

Once the command is executed, the necessary files will be downloaded and the container will start and will be accessible on port 8000.

There is also an option to setup Seafile with an SSL cert from Let’s Encrypt, but I didn’t do this because I already had a cert that I wanted to reuse.

Once the server is up, navigate to https://seafile.example.com/sys/settings/ and change SERVICE_URL and FILE_SERVER_URL to match your domain name.

Configure Apache 2

Adding a reverse proxy on Apache 2 will allow Apache 2 to route requests to port 8000, which will go to port 80 in the container. The main reference for this section is this: https://manual.seafile.com/deploy/deploy_with_apache/. This is done by adding a new site to Apache 2. The following configuration assumes that a LetsEncrypt SSL certificate is used.

cd /etc/apache2/sites-available
sudo nano seafile.conf

<IfModule mod_ssl.c>
<VirtualHost *:443>
    ServerName seafile.example.com
    ServerAdmin me@example.com

    RewriteEngine On

    # ModSecurity does not process requests.  There is a hard limit of 1 GB
    # with ModSecurity.
    SecRuleEngine Off

    # Whitelist internal IPs for mod_evasive
    DOSWhitelist 192.168.1.*

    <Location /media>
        Require all granted
    </Location>

    # Stuff for seafile server
    ProxyPass /seafhttp http://127.0.0.1:8082
    ProxyPassReverse /seafhttp http://127.0.0.1:8082
    RewriteRule ^/seafhttp - [QSA,L]

    # Stuff for seahub
    SetEnvIf Authorization "(.*)" HTTP_AUTHORIZATION=$1
    ProxyPreserveHost On
    ProxyPass / http://127.0.0.1:8000/
    ProxyPassReverse / http://127.0.0.1:8000/

    ErrorLog ${APACHE_LOG_DIR}/seafile_error.log
    CustomLog ${APACHE_LOG_DIR}/seafile_access.log combined

    Include /etc/letsencrypt/options-ssl-apache.conf

    SSLCertificateFile /path/to/cert/file.pem
    SSLCertificateKeyFile /path/to/cert/key/file.pem
</VirtualHost>
</IfModule>

I’ve added my local domain to the whitelist so that mod_evasive doesn’t think my local host is a threat.

There are two sets of ProxyPass and ProxyPassReverse lines. The reason for this is that Seafile is split up into Seahub and Seafile, where one uses port 8082 (by default), while the other uses 8000. These two lines are essentially what allows Apache to read requests from port 443 and pass them on to the container.

Once the file is saved, execute the following

sudo a2ensite seafile.conf
sudo service apache2 reload

Navigating to the site at the Apache 2 defined port should now load Seafile without having to go to port 8000.

Setup systemd

systemd allows the Docker container to be started up on boot, and be controlled like this:

sudo service seafile start/stop/restart

The configuration is as follows:

sudo nano /etc/systemd/system/seafile.service

[Unit]
Description=Seafile Server
After=network.target mysql.service

[Service]
Type=oneshot
ExecStart=/usr/bin/docker container start seafile
ExecStop=/usr/bin/docker container stop seafile
ExecReload=/usr/bin/docker container restart seafile
RemainAfterExit=yes
User=root
Group=root

[Install]
WantedBy=multi-user.target

Once the configuration is saved, shut off the existing running instance of Seafile and then let systemd start it up for you.

sudo systemctl daemon-reload
sudo service seafile start

At this point, Seafile should start automatically on boot, with Apache 2 doing a reverse proxy to the true Seafile port.

Setup e-mail (optional)

Optionally, Seafile can be configured to send e-mails. The reference is here: https://manual.seafile.com/config/sending_email/. The config file should be under:

/opt/seafile-data/seafile/conf/seahub_settings.py

I happen to be using Gmail SMTP, so at the end of the file I’ve added

EMAIL_USE_TLS = True
EMAIL_HOST = 'smtp.gmail.com'
EMAIL_HOST_USER = 'me@example.com'
EMAIL_HOST_PASSWORD = 'password'
EMAIL_PORT = 587
DEFAULT_FROM_EMAIL = EMAIL_HOST_USER
SERVER_EMAIL = EMAIL_HOST_USER

Backing up the image

After setting everything up, you’ll probably want to backup the Docker image. First, print out the list of running containers

sudo docker ps

Then take a snapshot of the current running state:

docker commit -p <container ID> <whatever you want to call this snapshot>

Running this command lists all the images and should show this newly saved image:

sudo docker images

I don’t have a private docker repository, so I saved the image to the disk and backed up the file by doing the below. Just make sure to actually move the image to an actual backup location.

sudo docker save -o <whatever path you want>/<filename>.tar <the name used earlier>

The image should be backed up now. To restore the image, it’s a matter of running this:

sudo docker load -i <path to the backup file>

Listing the images as above should show that the image was successfully loaded. Running this starts it back up again.

sudo docker run <container name>

Troubleshooting

Seafile logs can be helpful if it doesn’t start up correctly. These can be found at:

/opt/seafile-data/logs/seafile

There may be a need to ssh into the Docker container. The command for that is:

sudo docker exec -it seafile /bin/bash

Bash for Windows has been working pretty great for me until I needed to run Selenium Webdriver on it. I quickly learned that it wouldn’t work just work right out of the box, and the set up for it is quite convoluted.

You will need

• Bash for Windows
• Xming – This will actually allow the Window to appear on the screen
• geckodriver – Selenium now needs this to run Firefox

Bash setup

Install Firefox:

sudo apt-get install firefox

Export DISPLAY variable in ~/.bashrc. Just add this to the ~/.bashrc:

export DISPLAY=:0

Xming setup

Just download it from the link above and run it.

Download geckodriver

Download geckodriver from the link above and put in the /usr/bin/ folder in Bash for Windows.

Running Selenium

Here is a piece of sample Python code I used to setup the web browser. It will setup the browser, open Google, sit there for 10 seconds and then quit.  Make sure to have Xming running otherwise the browser isn’t going to start.

import time

from selenium import webdriver
from selenium.webdriver import DesiredCapabilities

def execute_with_retry(method, max_attempts):
    e = None
    for i in range(0, max_attempts):
        try:
            return method()
        except Exception as e:
            print(e)
            time.sleep(1)
    if e is not None:
        raise e

capabilities = DesiredCapabilities.FIREFOX
capabilities["marionette"] = True
firefox_bin = "/usr/bin/firefox"
browser = execute_with_retry(lambda: webdriver.Firefox(
    firefox_binary=firefox_bin, capabilities=capabilities), 10)

browser.get("https://www.google.com")

time.sleep(10)

browser.close()

• Self hosted installations allow greater control over what plugins the blog uses, what domain name to use, monetization, among other things
• The cost is the added time required to tweak to do the setup and customisation

After I chose to switch to the self-hosted installation, I faced the problem of having to migrate all of my content, site stats, and followers.  I also had to set up proper URL redirection to the new website.  Most of the guides I read did answer how to migrate content, and set up URL redirects, but it was not so clear as to how to migrate my site stats and followers.

Content Migration

For migrating content, the WordPress.com blog allows exporting blog content into a file to be imported into the new WordPress installation.  Steps 1 to 3 in this guide illustrate the process: http://www.wpbeginner.com/wp-tutorials/how-to-properly-move-your-blog-from-wordpress-com-to-wordpress-org/

Migrating Site Stats and Followers

Migrating site stats and followers requires installing the Jetpack addon and linking it to the WordPress.com account (so that the self-hosted WP and the existing WP blog are connected).  Afterward, a support thread has to be made on WordPress.com to ask WP staff to do the rest of the migration (there seems to be no other way).  I found the information on how to do this at this site: http://wordpress.stackexchange.com/questions/161633/migrating-stats-from-wordpress-com-blog-to-self-hosted-wordpress-org-blog.  The post I made for my own site migration is here.

Setting Up URL Redirection

This part costs money unfortunately.  I paid ~$17 CAD for this so that my own URL would be redirected for the new one.  If redirection isn’t required, then there is no need to buy anything.  The setup is pretty straight forward.  After buying the redirect, it is just a matter of setting the destination WordPress site to go to.  Reference link here: https://en.support.wordpress.com/site-redirect/

First, run:

sudo apt-get install dconf-editor

Navigate to:

org/gnome/desktop/applications/input-sources

Bold options are options that have been modified.  The interesting one is the row labeled “sources”.  Clear the entries of the IME’s that should not appear in the list and that should remove the IME’s that shouldn’t be appearing.

Setup

I pretty much used the instructions here: http://www.visualsvn.com/server/trac/.  I didn’t do step 8 written in the instructions, and it still worked for me.  However, what’s not setup is the Trac administrator account.  After the setup, I was able to access the Trac webapp and was also able to view the SVN logs, revisions, etc.  I initially got a message about “no changesets”, but after a commit to the repo, everything showed up on the webapp.

Setting up the admin account

While the above instructions allow you to login to Trac using the existing SVN credentials, that user is not considered an admin user.  All I had to do was run this command:

trac-admin /path/to/projenv permission add bob TRAC_ADMIN

After restarting the server, the admin tab was available in Trac.  Now Trac is available for use for bug tracking with an existing SVN repo!

Link: http://www.heerengandhi.com/2009/07/29/firefox-icon-went-missing-in-windows-7-heres-a-fix/comment-page-1/

You can configure the proofreading feature by going to My Account -> Edit Profile from the Dashboard.

Japadogs were at UBC today.  After wrestling with my impulses I decided that I wouldn’t buy one.  Even if they were incredibly good Japanese style hot dogs.