Improper planned updates will hurt entire organization

by Tseng H. K.
The article I read this week is “GitHub Says Database Issues Caused This Week’s Outage and Performance Problems” by Alex Williams.The article is saying that Github’s website had experienced outage and poor availability about a month ago. Company Github is a Social network for many programmers, it offers both private and public repository for many users. Team Github replaced old server to 3-node cluster in August, and new infrastructure is also designed based on new system. But the problem caused when high loads that Github never get. In result, some of private repositories were able to access to people who are not in the repository’s list for 7 minutes.

In the class, we have learned about entity and property. When we design database, we carefully need to think about what will be the data type and length. If we don’t plan well, it will most likely lose customers, and have problems soon or later. Having down time for web-site is very critical to many business, therefore, updates should be test very carefully. We also mentioned about it will be very costly when we try to change after design phase.

When we having updates, especially in live website. Developers need to spend time to test in every single way. I guess Github’s incident was caused by negligence of testing. When I do my projects in Java class last quarter, I always thought my project is bug-free. But, somehow the professor find a way to crash my project. Therefore, there are many unexpected things will happen, so we should test our projects very carefully over and over.

Citation:
Williams, A.(September 14, 2012) GitHub Says Database Issues Caused This Week’s Outage and Performance Problems Retrieved from http://techcrunch.com/2012/09/14/github-explains-this-weeks-outage-and-poor-performance/

3 thoughts on “Improper planned updates will hurt entire organization

  • October 7, 2012 at 5:31 pm
    Permalink

    I thought that the article you chose was very interesting. It shows a clear example of the importance of database design and the consequences of poor planning. The fact that there was downtime for GitHub must have hurt its business and costs thousands of dollars to fix. With careful planning, all of this could have been avoided to begin with. This shows how it is necessary to take the time to create a proper database design as not to cause problems in the future.

  • October 7, 2012 at 6:59 pm
    Permalink

    As much as I agree that careful planning can help reduce the consequences of problems in the future, I think maintenance is also an important factor. There is only so much planning an organization can do at first but they’re not going to catch every little mistake. Even to this day many successful sites still have issues with their sites due to high loads but these sites quickly take care of any issues that may occur. Careful planning is a great way to help prevent problems in the future but I don’t think it’s the only solution. I think fast and reliable maintenance is another.

  • October 7, 2012 at 9:18 pm
    Permalink

    It’s interesting to know that a company like GitHub implemented an untested solution. I personally have a public GitHub account and know many programmers that have private accounts. Their accounts host proprietary code that they probably do not wish to share with the public. This article reinforces the importance of testing before deployment…no matter how big or small the changes are.

Comments are closed.