Current time: 08-09-2020, 01:26 PM Hello There, Guest! (LoginRegister)

Post Reply 
 
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
EC2 instance size: t3.nano seems to work...?
06-17-2020, 09:23 AM
Post: #1
EC2 instance size: t3.nano seems to work...?
Everyone,

I'm exploring downsizing our instances to the smallest (cheapest) possible EC2 size: t3.nano. Where does this break?

I've bombarded our Virginia queue with about a thousand tests and they're dutifully completing at about 1 per minute.

I've read this https://www.webpagetest.org/forums/showt...?tid=13460
...and this https://www.webpagetest.org/forums/showt...?tid=12007 ... but sadly the link "Which EC2 instance type should be used for WebPageTest Agents?" is broken.

Our use case is very minimal:
+ 1 test every min
+ 1 run
+ First view only
+ I'm using "google.com" for testing (arbitrary...I'll probably try cnn.com, etc. after gathering some statistics)
+ IE is not our primary use case

Any thoughts, reading, data would be excellent.

Thanks.
Find all posts by this user
Quote this message in a reply
06-18-2020, 12:24 AM
Post: #2
RE: EC2 instance size: t3.nano seems to work...?
Oh, it will "work" but the results will be worthless. You want a c4.large (or something along those lines) for consistent results.

The t* line are burstable CPU performance so the first few tests will run fast and then the CPU will be throttled and the rest of the tests will be MUCH slower (and every now and then burst faster).

Outside of never using the t* series, you need to make sure you have enough CPU horsepower to be representative of your visitors. Using too slow of a VM (too few VCPUs) will cause the JS to run much slower.
Visit this user's website Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)