Closed ethanmye-rs closed 3 months ago
Hi @ethanmye-rs Thanks for reporting! and apologies for the delay in responding - we just go to clean up old issues etc..
Those kernel parameters are ephemeral and get reset after a reboot when set using sysctl
. Should you want these to be persisted across reboots you need to set them in /etc/sysctl.conf
, i.e:
On the host:
sudo tee -a /etc/sysctl.conf > /dev/null <<EOT
vm.max_map_count=262144
vm.swappiness=0
net.ipv4.tcp_retries2=5
fs.file-max=1048576
EOT
sudo sysctl -p
For the juju units when creating the model:
cat <<EOF > cloudinit-userdata.yaml
cloudinit-userdata: |
postruncmd:
- [ 'echo', 'vm.max_map_count=262144', '>>', '/etc/sysctl.conf' ]
- [ 'echo', 'vm.swappiness=0', '>>', '/etc/sysctl.conf' ]
- [ 'echo', 'net.ipv4.tcp_retries2=5', '>>', '/etc/sysctl.conf' ]
- [ 'echo', 'fs.file-max=1048576', '>>', '/etc/sysctl.conf' ]
- [ 'sysctl', '-p' ]
EOF
juju model-config --file=./cloudinit-userdata.yaml
We had updated the README - but looks like we missed the discourse docs, which I updated now.
Please reopen this issue if it persists.
Steps to reproduce
sudo sysctl -w vm.max_map_count=262144 vm.swappiness=0 net.ipv4.tcp_retries2=5
as they were reset after rebootsudo sysctl -a | grep -E 'swappiness|max_map_count|tcp_retries2'
I've exposed the opensearch service, but I don't think that matters here.
Expected behavior
The tutorial (and my experience) says this should work OK. The other parameters are successfully set. A reboot of the container results in the same error message.
Actual behavior
Juju does not recognize net.ipv4.tcp_retries2 as being reset. I've tried rebooting the containers without any luck. I suspect this is more of a juju problem than an opensearch problem, but it would be nice to figure out how to unstick juju.
Versions
opensearch active 1 opensearch 2/edge 26 yes
Operating system:22.04
Happy to provide any logs required.