Testnet processing accounts keeps going back to zero percent

post by ramdio123 on May 1, 2023

Hi everyone!!!

I tried to run testnet and when it comes to processing accounts progress it always stops at 30% or 40% and returns to 0%

What could be the possible solutions?!

post by fabrice on May 1, 2023

I’m guessing either your machine is under-powered (not enough RAM or using an HDD instead of SSD).

Or, your algorand software is not up to date.

Please see Node Troubleshooting - Algorand Developer Portal for a detailed troubleshooting guide.

post by ramdio123 on May 1, 2023

my co-interns has only HDD installed on their laptop but was able to set up testnet successfully, I am the only one who has 4GB of RAM

post by fabrice on May 1, 2023

Your co-intern has 8GB? And you have 4GB?

In that case, this is the most likely issue.

If you need to develop on Algorand on your computer, you can either use algokit to create a private network (which should not have this issue as it’ll have way fewer transactions) or use a free API service, if you really need TestNet.

I would recommend the former.

post by ramdio123 on May 1, 2023

They indeed have an 8GB ram in their laptops

And we are now in the lesson where we need to set up testnet

So I plan to buy additional Ram this week

post by pavel on May 2, 2023

@ramdio123 check node.log log file for additional messages around the time it flips from 40% to 0%. Try warning, error, catch terms for grepping. The processing loops restarts in case of errors - it could be a networking issue, decoding error or as Fabrice pointed out a memory issue.

post by ramdio123 on May 3, 2023

Is it also fine to install 120gb SSD for better performance of my laptop?

post by ramdio123 on May 3, 2023

I replaced my HDD with 120gb SSD this is the result

I still have my 4gb ram in my laptop

post by pavel on Aug 11, 2023

Sorry for the late reply. Memory constraint could be an issue during the fast catchup especially running sandbox (docker) on Windows. We are working on some memory optimizations and I created a ticket to ensure this 4GB use case supported.