Sihai network

What's wrong with alicloud's failure? What is the cause of alicloud's failure?

On the afternoon of June 27, many netizens reported that Alibaba cloud platform had access failure in Weibo. Many products have broken down. What's the matter? Let's get to know.

We learned from the Alibaba cloud official website announcement that the start time of the fault is around 16:21 on June 27, 2018. The main reason is that some control functions of Alibaba cloud official website and some functions of MQ, NAS, OSS and other products have access exceptions. However, most functions have now returned to normal.

Later, Alibaba cloud published a fault description, in which Alibaba cloud said: there is no excuse for this fault, we can't and shouldn't make such a mistake! We will seriously improve the automatic operation and maintenance technology and release the verification process, awe every line of code, awe every payment.

The following is the original alicloud fault description:

In the afternoon of June 27th, a mistake in our operation and maintenance resulted in some customers visiting Ali's official website console and using some of its products, which caused a lot of problems. The fault started at around 16:21 on June 27, 2018, Beijing time, and began to recover at 16:50.

After the emergency technical restoration, the causes of the failure are as follows:

In the afternoon of that day, the engineer team carried out a change verification operation in a new automatic operation and maintenance function. This function has no problem in the verification of test environment. When it is online to the automatic operation and maintenance system, an unknown code bug is triggered, and some internal IP is disabled by the error code, resulting in the access link of some products is blocked. After the follow-up manual intervention, the engineer team quickly recovered the problem.

Affected locations include Alibaba cloud official website console, MQ, NAS, OSS and other product functions. There is no excuse for this failure, we can't and shouldn't make such a mistake! We will seriously improve the automatic operation and maintenance technology and release the verification process, awe every line of code, awe every payment.