- Newest
- Most votes
- Most comments
I am currently facing exactly the same problem. Everything seems to be running fine and all of sudden the url is inaccessible. Only fix is to stop the Wordpress instance and start it again. Looking into the metrics I see it's CPU is around 2-3% and all of sudden bursts tot 20%+. The website has very little traffic. I'm getting pretty frustrated here as after reading this thread, I get the idea its an ongoing issue with Lightsail since 2017... Now looking into options of moving from AWS as I can't keep having this.
Unfortunately, at least according to the latest documentation, you're running into a limitation of LightSail instance types and functionality.
- For the CPU - If you can either add WP caching, or DB indexes, or something like that, you may be able to reduce the amount of load on the instance, which would hopefully keep the overall amount of CPU from bursting for too long, and subsequently hamstringing your server to the point that it is unusable. Newer EC2 instances have a thing called unlimited burst, which, for an extra charge, allows your instance to sustain higher activity even after you run out of credits, but it doesn't look like those have found their way to the LightSail platform yet.
- For the plan upgrades - Yeah, making you snapshot and upgrade your instance instead of offering sliders or some other automated upgrade version is less than desirable. Unfortunately, this is another limitation imposed by the tech it's using under the hood. That's not to say that it can't be done, just that they haven't implemented anything to address it yet.
I wish I had a better response, and I hope they address this soon. The other option is to move away from LightSail and use AWS EC2, and choose an instance type that offers more consistent performance.
Thanks for your response! I came to a more or less similar conclusion. It's just the thing that lightsale as a product offers "click & go" easy to use and setup Software as a Service with pre-configured stacks (WordPress, Prestashop, etc.). However, if it can't even uphold that for a simple use case at a 3.5$ plan, it should not be offered in the first place IMHO. I had a couple of instances set up for my students to play with and create online shops and websites and after 3 weeks almost all machines were not usable anymore, quite a pain in the lower back.
It's also a bit strange, that the 3.5$ plan with 512MB and 1 Core is not enough and bringing the machine to a halt so quickly. I've dealt with webservers quite some time and that hardware usually should be sufficient for small 3-tier web applications like WP, etc. ?
So my conclusion would be to only use the 3.5$ (and 5$) plan to do some fiddling and just "checking out" with a web stack.
I have same issue. My instance keeps going down every two or three days. I don't see any burst in the metric graph. https://imgur.com/ewaGlJU. The dips of the remaining burst capacity shown in this chart were caused by rebooting.
I have same issue with the 1G plan (WordPress installed).
The CPU usage stays under sustainable zone, and remaining burst capacity stays at 100%. Every two or three days the instance just downs. I don't see the burst at all, and memory usage is less 50%.
Still an issue. I'm not sure the point of Lightsail. I thought it would be reliable.
Relevant content
- asked 2 years ago
- asked a year ago
- asked a year ago
- AWS OFFICIALUpdated 3 months ago
- AWS OFFICIALUpdated 3 months ago
- AWS OFFICIALUpdated 4 months ago
- AWS OFFICIALUpdated 4 months ago
did you fix this eventually?