Skip to content
Snippets Groups Projects

Update readme for use with ood

Merged John-Paul Robinson requested to merge feat-update-readme into master
1 file
+ 38
17
Compare changes
  • Side-by-side
  • Inline
+ 38
17
Project to provision an OpenHPC cluster via Vagrant using the
Project to provision an [OpenHPC](https://openhpc.community/) + [Open OnDemand](https://openondemand.org/) cluster via Vagrant using the
CRI_XCBC (XSEDE basic cluster) Ansible provisioning framework.
CRI_XCBC (XSEDE basic cluster) Ansible provisioning framework.
The Vagrantfile takes inspiration from the [vagrantcluster](https://github.com/cluening/vagrantcluster)
The Vagrantfile takes inspiration from the [vagrantcluster](https://github.com/cluening/vagrantcluster)
project but is oriented toward deploying only a master node
project but is oriented toward deploying only a master node
and using standard OHPC tools to provision the cluster, and
and using standard OHPC tools to provision the cluster, and
therfore favors the CRI_XCBC approach to ansible scripts just
therfore favors the CRI_XCBC approach to ansible scripts just
for the master.
for the master.
The Vagrantfile is stripped to the core (rather that carry all
The Vagrantfile is stripped to the core (rather that carry all
the cruft of a vagrant init). It leverages work from a
the cruft of a vagrant init). It leverages work from a
[pilot project](https://gitlab.rc.uab.edu/ravi89/ohpc_vagrant)
[pilot project](https://gitlab.rc.uab.edu/ravi89/ohpc_vagrant)
(primaryly the development of an updated centos 7.5 image)
(primaryly the development of an updated centos 7.5 image)
but prefers a clean repo slate.
but prefers a clean repo slate.
## Project Setup
## Project Setup
After cloning this project you need to initialize the submodule
Clone this project recursively to get the correct version for the
from with in the git repo
CRI_XSEDE submodule to build the OpenHPC(ohpc) and Open OnDemand (ood) nodes
```
```
git submodule init
git clone --recursive https://gitlab.rc.uab.edu/jpr/ohpc_vagrant.git
git submodule update
```
```
Alternatively you can provide the `--recurse-submodules` command
during the initial clone.
## Cluster Setup
## Cluster Setup
After setting up the project above create your single node OpenHPC
After setting up the project above create your single node OpenHPC
cluster with vagrant:
cluster with vagrant:
```
```
vagrant up
vagrant up ohpc
```
```
The ansible config will bring the master node to the point where its
The ansible config will bring the master node to the point where its
@@ -43,12 +39,12 @@ Create node c0 (choose whatever name makes sense, c0 matches the config):
@@ -43,12 +39,12 @@ Create node c0 (choose whatever name makes sense, c0 matches the config):
compute_create c0
compute_create c0
```
```
When prompted start node c0:
When prompted start compute node c0:
```
```
compute_start c0
compute_start c0
```
```
If you want to stop the node:
If you want to stop the compute node:
```
```
compute_stop c0
compute_stop c0
```
```
@@ -65,7 +61,7 @@ ipxe.iso in compute_create to match your local environment.
@@ -65,7 +61,7 @@ ipxe.iso in compute_create to match your local environment.
## Cluster Check
## Cluster Check
After the `vagrant up` completes you can can log into the cluster with `vagrant ssh`.
After the `vagrant up ohpc` completes you can can log into the cluster with `vagrant ssh ohpc`.
To confirm the system is operational run `sinfo` and you should see the following text:
To confirm the system is operational run `sinfo` and you should see the following text:
```
```
@@ -82,3 +78,28 @@ srun hostname
@@ -82,3 +78,28 @@ srun hostname
This should return the name `c0`.
This should return the name `c0`.
With these tests confirmed you have a working OpenHPC cluster running slurm.
With these tests confirmed you have a working OpenHPC cluster running slurm.
 
 
## Boot the Open OnDemand node
 
 
A primary function of this project is to provide a dev/test cluster for working
 
with Open OnDemand. After the cluster is up boot the ood node with:
 
```
 
vagrant up ood
 
```
 
 
This will provision the node and near the end of the provisioning provide several
 
sudo commands that need to be run on the ohpc node to register the ood node
 
with the cluster, ensuring data synchronization and slurm work.
 
 
After the node is provisioned (or booted) you need to work around mount issue
 
with NFS mounts and issue the `mount -a` command on the ood node:
 
```
 
vagrant ssh ood -c "sudo mount -a"
 
```
 
 
After this point you can connect to the web ui of the ood node, typically via
 
(the port mapping may change in your local vagrant env):
 
 
http://localhost/8080
 
 
The default user name and password for the web UI is 'vagrant'.
Loading