Skip to content
This repository has been archived by the owner on Nov 7, 2018. It is now read-only.

max file descriptors is too low on AWS EKS #215

Open
ksemaev opened this issue Aug 17, 2018 · 7 comments
Open

max file descriptors is too low on AWS EKS #215

ksemaev opened this issue Aug 17, 2018 · 7 comments

Comments

@ksemaev
Copy link

ksemaev commented Aug 17, 2018

So debugged it from very begining on my node:

systcl -a | grep vm.max_map_count
>vm.max_map_count = 262144
cat /proc/sys/fs/file-max
>1636119
ulimit -Hn
>4096

docker pull quay.io/pires/docker-elasticsearch-kubernetes:6.2.3
systcl -a | grep vm.max_map_count
>vm.max_map_count = 262144
cat /proc/sys/fs/file-max
>1636119
ulimit -Hn
>4096

/run.sh
>...
>ERROR: [1] bootstrap checks failed
>[1]: max file descriptors [4096] for elasticsearch process is too low, increase to at least [65536]

I believe that you should specify in prerequsites ulimit setup

@ksemaev
Copy link
Author

ksemaev commented Aug 17, 2018

In my case it was defaut docker setup on EKS, had to do on node (or add to AMI):
sed -i -e 's/1024:4096/65536:65536/g' /etc/sysconfig/docker

@pires
Copy link
Owner

pires commented Aug 20, 2018

Thanks for the suggestion. I'll leave this here for future reference.

@gageorsburn
Copy link

Out of all of the cat memes on the internet, there exists only one issue for this and it was extremely difficult to find.

@pires
Copy link
Owner

pires commented Aug 31, 2018

Hopefully, this tweet will help.

@slavaaaaaaaaaa
Copy link

@ksemaev just wondering, what were the symptoms of this?

@ksemaev
Copy link
Author

ksemaev commented Sep 7, 2018

@smaslennikov elastic was starting log with that warning message

@maauso
Copy link

maauso commented Oct 16, 2018

Hi,

we found the same problem using the last amazon ami (ami-0440e4f6b9713faf6), but if we use the previous one (ami-0b2ae3c6bda8b5c06), we didn't find this problem.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

5 participants