About Blog Migration

Posted on Views: Waline: Word count in article: 345 Reading time ≈ 1 mins.
The author originally built a blog on a Tencent Cloud server based on Django. Due to the server's expiration and a change in its intended use, the author decided to migrate. This time, Hexo+Next was used to generate a static website and host it on Alibaba Cloud Object Storage Service, saving 85% in costs. The author also shares the writing-to-publication process. However, some data was lost during the migration.

My blog was previously hosted on a Tencent Cloud server, which was developed based on a secondary development of a Django open-source project.

The server will expire in November this year. Additionally, I recently need to use that server for other purposes. So, I’m considering migrating the server.

One of the main reasons I chose to develop the blog with Django was that I hoped to develop my own blog and also get some hands - on practice (I wasn’t very proficient in Django development at that time).

However, in the three years of writing the blog, I’ve found that a static website hosting can meet most of the requirements for a blog website. So this time, I used Hexo+Next to generate a static website and then hosted it on Alibaba Cloud’s Object Storage Service. This is much cheaper in terms of cost compared to buying a server before, with an estimated cost savings of 85%.

For me, apart from the blog’s display function itself, a good writing experience is also very important. Here is my current process from writing to publication:

  1. First, I write locally using Obsidian, mostly in Markdown format.
  2. Through the Git plugin, local changes are automatically synced to the Obsidian repository on GitHub every two minutes.
  3. There is a Hexo repository that syncs the content from the Obsidian repository at 6 am every morning and then redeploys it to Alibaba Cloud OSS.

Doing this has two advantages: First, it separates the writing process from the blog publication, allowing me to focus on writing and leave the rest to the pre - defined pipeline. Second, all the blog posts are Markdown files, and there is a copy both locally and on GitHub, so I don’t have to worry about losing them. It’s also very convenient for doing other things, such as publishing the blog to WeChat Official Accounts based on these Markdown files.

The most regrettable thing about this blog migration is that I lost the previous reading volume data and readers’ comment data because they couldn’t be migrated.