Out of Memory error and process killed while Installing esp32 core in EC2 Instance (Free Tier Ubuntu)

[DevsBotIOT]

Describe the problem

I was trying to install esp32 core arduino-cli in EC2 Free (t1.micro Instance). By using this https://lindevs.com/install-arduino-cli-on-ubuntu I have installed Arduino-cli. After that I issued the below commands.

arduino-cli config init
sudo nano /home/ubuntu/.arduino15/arduino-cli.yaml (updated additional-urls ["https://espressif.github.io/arduino-esp32/package_esp32_index.json"])
arduino-cli core update-index --config-file .arduino15/arduino-cli.yaml
arduino-cli config dump =[shows additional-urls esp32 link]
arduino-cli core install esp32:esp32

At the last command, starting seems to be running smooth and at the final esp32 installation it hangs for few minutes and got a message killed and the installation stops. I tried many times and same result. I Checked the /var/log/kern.log and I found this

/var/log/kern.log
May 9 10:06:53 ip-172-31-5-99 kernel: [7260893.611782] Out of memory: Killed process 402210 (arduino-cli) total-vm:1399256kB, anon-rss:682408kB, file-rss:0kB, shmem-rss:0kB, UID:1000 pgtables:1448kB oom_score_adj:0

To reproduce

arduino-cli config init
sudo nano /home/ubuntu/.arduino15/arduino-cli.yaml (updated additional-urls ["https://espressif.github.io/arduino-esp32/package_esp32_index.json"])
arduino-cli core update-index --config-file .arduino15/arduino-cli.yaml
arduino-cli config dump =[shows additional-urls esp32 link]
arduino-cli core install esp32:esp32

Expected behavior

esp32 core must be installed after the command.

Arduino CLI version

0.32.2

Operating system

Linux

Operating system version

22.04.01

Additional context

No response

Issue checklist

• I searched for previous reports in the issue tracker
• I verified the problem still occurs when using the nightly build
• My report contains all necessary details

[umbynos]

Hi @DevsBotIOT, this seems related to #1568 and #1059. And the high memory usage should be solved by #981. To me it seems that you are trying to install a 3rd party core to a low resource system. Unfortunately I think there is nothing we can do.. But I can suggest to upgrade to an ec2 instance with more RAM

[umbynos]

Hi again @DevsBotIOT, I took some time today to investigate your problem and after a bit of research on how to perform memory profiling in golang I got some results using pprof (:exploding_head: ):
I run (on a 64bit system) ./arduino-cli core install esp32:esp32 --additional-urls https://espressif.github.io/arduino-esp32/package_esp32_index.json:

Apparently, the functions allocating the majority of memory are the one bound to the file-system operations: the archive extraction. I would like to keep this issue open, so that in the future we can investigate if something could be done in order to optimize this.