add spot instance support and pricing
This commit is contained in:
Robert Wilkins III
Gabriel Luiz Freitas Almeida
parent
6a59d64a0d
commit
dada24a5e6
4 changed files with 221 additions and 6 deletions
|
|
@ -1,7 +1,30 @@
|
|||
# Running Langflow from a new Google Cloud project
|
||||
This guide will help you set up a Langflow Dev VM in a Google Cloud Platform project using Google Cloud Shell.
|
||||
> When cloudshell opens, select **Trust repo**. Some gcloud commands do not run in ephemeral cloudshell.
|
||||
# Run Langflow from a New Google Cloud Project
|
||||
|
||||
This guide will help you set up a Langflow development VM in a Google Cloud Platform project using Google Cloud Shell.
|
||||
|
||||
> **Note**: When Cloud Shell opens, be sure to select **Trust repo**. Some `gcloud` commands might not run in an ephemeral Cloud Shell environment.
|
||||
|
||||
|
||||
## Standard VM
|
||||
[](https://console.cloud.google.com/cloudshell/open?git_repo=https://github.com/genome21/langflow&working_dir=scripts&shellonly=true&tutorial=walkthroughtutorial.md)
|
||||
|
||||
This script sets up a Debian-based VM with the Langflow package, Nginx, and the necessary configurations to run the Langflow Dev environment.
|
||||
This script sets up a Debian-based VM with the Langflow package, Nginx, and the necessary configurations to run the Langflow Dev environment.
|
||||
<hr>
|
||||
|
||||
## Spot/Preemptible Instance
|
||||
|
||||
[](https://console.cloud.google.com/cloudshell/open?git_repo=https://github.com/genome21/langflow&working_dir=scripts&shellonly=true&tutorial=walkthroughtutorial.md)
|
||||
|
||||
When running as a [spot (preemptible) instance](https://cloud.google.com/compute/docs/instances/preemptible), the code and VM will behave the same way as in a regular instance, executing the startup script to configure the environment, install necessary dependencies, and run the Langflow application. However, **due to the nature of spot instances, the VM may be terminated at any time if Google Cloud needs to reclaim the resources**. This makes spot instances suitable for fault-tolerant, stateless, or interruptible workloads that can handle unexpected terminations and restarts.
|
||||
|
||||
## Pricing
|
||||
<br>
|
||||
|
||||
| Component | Regular Cost (Hourly) | Regular Cost (Monthly) | Spot/Preemptible Cost (Hourly) | Spot/Preemptible Cost (Monthly) | Notes |
|
||||
| -------------- | --------------------- | ---------------------- | ------------------------------ | ------------------------------- | ----- |
|
||||
| 100 GB Disk | - | $10/month | - | $10/month | Disk cost remains the same for both regular and Spot/Preemptible VMs |
|
||||
| VM (n1-standard-4) | $0.15/hr | ~$108/month | ~$0.04/hr | ~$29/month | The VM cost can be significantly reduced using a Spot/Preemptible instance |
|
||||
| **Total** | **$0.15/hr** | **~$118/month** | **~$0.04/hr** | **~$39/month** | Total costs for running the VM and disk 24/7 for an entire month |
|
||||
|
||||
> For a more accurate breakdown of costs, please use the [**GCP Pricing Calculator**](https://cloud.google.com/products/calculator)
|
||||
|
||||
|
|
|
|||
101
scripts/deploy_langflow_gcp_spot.sh
Normal file
101
scripts/deploy_langflow_gcp_spot.sh
Normal file
|
|
@ -0,0 +1,101 @@
|
|||
# Set the VM, image, and networking configuration
|
||||
VM_NAME="langflow-dev"
|
||||
IMAGE_FAMILY="debian-11"
|
||||
IMAGE_PROJECT="debian-cloud"
|
||||
BOOT_DISK_SIZE="100GB"
|
||||
ZONE="us-central1-a"
|
||||
REGION="us-central1"
|
||||
VPC_NAME="default"
|
||||
SUBNET_NAME="default"
|
||||
SUBNET_RANGE="10.128.0.0/20"
|
||||
NAT_GATEWAY_NAME="nat-gateway"
|
||||
CLOUD_ROUTER_NAME="nat-client"
|
||||
|
||||
# Set the GCP project's compute region
|
||||
gcloud config set compute/region $REGION
|
||||
|
||||
# Check if the VPC exists, and create it if not
|
||||
vpc_exists=$(gcloud compute networks list --filter="name=$VPC_NAME" --format="value(name)")
|
||||
if [[ -z "$vpc_exists" ]]; then
|
||||
gcloud compute networks create $VPC_NAME --subnet-mode=custom
|
||||
fi
|
||||
|
||||
# Check if the subnet exists, and create it if not
|
||||
subnet_exists=$(gcloud compute networks subnets list --filter="name=$SUBNET_NAME AND region=$REGION" --format="value(name)")
|
||||
if [[ -z "$subnet_exists" ]]; then
|
||||
gcloud compute networks subnets create $SUBNET_NAME --network=$VPC_NAME --region=$REGION --range=$SUBNET_RANGE
|
||||
fi
|
||||
|
||||
# Create a firewall rule to allow TCP port 8080 for all instances in the VPC
|
||||
firewall_8080_exists=$(gcloud compute firewall-rules list --filter="name=allow-tcp-8080" --format="value(name)")
|
||||
if [[ -z "$firewall_8080_exists" ]]; then
|
||||
gcloud compute firewall-rules create allow-tcp-8080 --network $VPC_NAME --allow tcp:8080 --source-ranges 0.0.0.0/0 --direction INGRESS
|
||||
fi
|
||||
|
||||
# Create a firewall rule to allow IAP traffic
|
||||
firewall_iap_exists=$(gcloud compute firewall-rules list --filter="name=allow-iap" --format="value(name)")
|
||||
if [[ -z "$firewall_iap_exists" ]]; then
|
||||
gcloud compute firewall-rules create allow-iap --network $VPC_NAME --allow tcp:80,tcp:443 --source-ranges 35.235.240.0/20 --direction INGRESS
|
||||
fi
|
||||
|
||||
# Create the Cloud Router and NAT Gateway
|
||||
cloud_router_exists=$(gcloud compute routers list --filter="name=$CLOUD_ROUTER_NAME" --format="value(name)")
|
||||
if [[ -z "$cloud_router_exists" ]]; then
|
||||
gcloud compute routers create $CLOUD_ROUTER_NAME --network $VPC_NAME --region $REGION
|
||||
fi
|
||||
|
||||
nat_exists=$(gcloud compute routers list --filter="name=$CLOUD_ROUTER_NAME" --format="value(nats.name)")
|
||||
if [[ -z "$nat_exists" ]]; then
|
||||
gcloud compute routers nats create $NAT_GATEWAY_NAME --router $CLOUD_ROUTER_NAME --auto-allocate-nat-external-ips --nat-all-subnet-ip-ranges --enable-logging --region $REGION
|
||||
fi
|
||||
|
||||
# Define the startup script as a multiline Bash here-doc
|
||||
STARTUP_SCRIPT=$(cat <<'EOF'
|
||||
#!/bin/bash
|
||||
|
||||
# Update and upgrade the system
|
||||
apt -y update
|
||||
apt -y upgrade
|
||||
|
||||
# Install Python 3 pip, Langflow, and Nginx
|
||||
apt -y install python3-pip
|
||||
pip install langflow
|
||||
apt-get -y install nginx
|
||||
|
||||
# Configure Nginx for Langflow
|
||||
touch /etc/nginx/sites-available/langflow-app
|
||||
echo "server {
|
||||
listen 0.0.0.0:8080;
|
||||
|
||||
location / {
|
||||
proxy_pass http://127.0.0.1:7860;
|
||||
proxy_set_header Host "\$host";
|
||||
proxy_set_header X-Real-IP "\$remote_addr";
|
||||
proxy_set_header X-Forwarded-For "\$proxy_add_x_forwarded_for";
|
||||
}
|
||||
}" >> /etc/nginx/sites-available/langflow-app
|
||||
ln -s /etc/nginx/sites-available/langflow-app /etc/nginx/sites-enabled/
|
||||
sudo nginx -t
|
||||
sudo systemctl restart nginx
|
||||
langflow
|
||||
EOF
|
||||
)
|
||||
|
||||
# Create a temporary file to store the startup script
|
||||
tempfile=$(mktemp)
|
||||
echo "$STARTUP_SCRIPT" > $tempfile
|
||||
|
||||
# Create the VM instance with the specified configuration and startup script
|
||||
gcloud compute instances create $VM_NAME \
|
||||
--image-family $IMAGE_FAMILY \
|
||||
--image-project $IMAGE_PROJECT \
|
||||
--boot-disk-size $BOOT_DISK_SIZE \
|
||||
--machine-type=n1-standard-4 \
|
||||
--metadata-from-file startup-script=$tempfile \
|
||||
--zone $ZONE \
|
||||
--network $VPC_NAME \
|
||||
--subnet $SUBNET_NAME \
|
||||
-preemptible
|
||||
|
||||
# Remove the temporary file after the VM is created
|
||||
rm $tempfile
|
||||
|
|
@ -39,7 +39,8 @@ The script will:
|
|||
3. Create a Compute Engine VM instance with the specified configuration and startup script
|
||||
4. Configure Nginx to serve Langflow on TCP port 8080
|
||||
|
||||
> <walkthrough-pin-section-icon></walkthrough-pin-section-icon> The process may take approximately 30 minutes to complete. Rest assured that progress is being made, and you'll be able to proceed once the process is finished.
|
||||
<walkthrough-pin-section-icon></walkthrough-pin-section-icon>
|
||||
> The process may take approximately 30 minutes to complete. Rest assured that progress is being made, and you'll be able to proceed once the process is finished.
|
||||
|
||||
In the next step, you'll learn how to connect to the Langflow VM.
|
||||
|
||||
|
|
@ -75,7 +76,9 @@ gcloud compute routers delete nat-client --region us-central1 --quiet
|
|||
The following network settings and services are used during this walkthrough. If you plan to continue using the project after the walkthrough, you may keep these configurations in place.
|
||||
|
||||
However, if you decide to remove them after completing the walkthrough, you can use the following gcloud commands:
|
||||
> <walkthrough-pin-section-icon></walkthrough-pin-section-icon> These commands will delete the firewall rules and network configurations created during the walkthrough. Make sure to run them only if you no longer need these settings.
|
||||
|
||||
<walkthrough-pin-section-icon></walkthrough-pin-section-icon>
|
||||
> These commands will delete the firewall rules and network configurations created during the walkthrough. Make sure to run them only if you no longer need these settings.
|
||||
|
||||
```
|
||||
gcloud compute firewall-rules delete allow-tcp-8080 --quiet
|
||||
|
|
|
|||
88
scripts/walkthroughtutorial_spot.md
Normal file
88
scripts/walkthroughtutorial_spot.md
Normal file
|
|
@ -0,0 +1,88 @@
|
|||
# Deploy Langflow on Google Cloud Platform
|
||||
|
||||
**Duration**: 45 minutes
|
||||
**Author**: [Robert Wilkins III](https://www.linkedin.com/in/robertwilkinsiii)
|
||||
|
||||
## Introduction
|
||||
|
||||
In this tutorial, you will learn how to deploy Langflow on [Google Cloud Platform](https://cloud.google.com/) (GCP) using Google Cloud Shell.
|
||||
|
||||
This tutorial assumes you have a GCP account and basic knowledge of Google Cloud Shell. If you're not familiar with Cloud Shell, you can review the [Cloud Shell documentation](https://cloud.google.com/shell/docs).
|
||||
|
||||
## Set up your environment
|
||||
|
||||
Before you start, make sure you have the following prerequisites:
|
||||
|
||||
- A GCP account with the necessary permissions to create resources
|
||||
- A project on GCP where you want to deploy Langflow
|
||||
|
||||
[**Select your GCP project**]<walkthrough-project-setup
|
||||
billing="true"
|
||||
apis="compute.googleapis.com,container.googleapis.com">
|
||||
</walkthrough-project-setup>
|
||||
|
||||
|
||||
In the next step, you'll configure the GCP environment and deploy Langflow.
|
||||
|
||||
## Configure the GCP environment and deploy Langflow
|
||||
Run the deploy_langflow_gcp_spot.sh script to configure the GCP environment and deploy Langflow:
|
||||
|
||||
```sh
|
||||
gcloud config set project <walkthrough-project-id/>
|
||||
bash ./deploy_langflow_gcp.sh
|
||||
```
|
||||
|
||||
The script will:
|
||||
|
||||
1. Check if the required resources (VPC, subnet, firewall rules, and Cloud Router) exist and create them if needed
|
||||
2. Create a startup script to install Python, Langflow, and Nginx
|
||||
3. Create a Compute Engine VM instance with the specified configuration and startup script
|
||||
4. Configure Nginx to serve Langflow on TCP port 8080
|
||||
|
||||
> <walkthrough-pin-section-icon></walkthrough-pin-section-icon> The process may take approximately 30 minutes to complete. Rest assured that progress is being made, and you'll be able to proceed once the process is finished.
|
||||
|
||||
In the next step, you'll learn how to connect to the Langflow VM.
|
||||
|
||||
## Connect to the Langflow VM
|
||||
To connect to your new Langflow VM, follow these steps:
|
||||
|
||||
1. Navigate to the [VM instances](https://console.cloud.google.com/compute/instances) page and click on the external IP for your VM. Make sure to use HTTP and set the port to 8080
|
||||
<br>**or**
|
||||
3. Run the following command to display the URL for your Langflow environment:
|
||||
```bash
|
||||
export LANGFLOW_IP=$(gcloud compute instances list --filter="NAME=langflow-dev" --format="value(EXTERNAL_IP)")
|
||||
|
||||
echo http://$LANGFLOW_IP:8080
|
||||
```
|
||||
|
||||
4. Click on the Langflow URL in cloudshell to be greeted by the Langflow Dev environment
|
||||
|
||||
Congratulations! You have successfully deployed Langflow on Google Cloud Platform.
|
||||
|
||||
<walkthrough-conclusion-trophy></walkthrough-conclusion-trophy>
|
||||
|
||||
## Cleanup
|
||||
If you want to remove the resources created during this tutorial, you can use the following commands:
|
||||
|
||||
```sql
|
||||
gcloud compute instances delete langflow-dev --zone us-central1-a --quiet
|
||||
|
||||
gcloud compute routers nats delete nat-gateway --router nat-client --region us-central1 --quiet
|
||||
|
||||
gcloud compute routers delete nat-client --region us-central1 --quiet
|
||||
|
||||
```
|
||||
The following network settings and services are used during this walkthrough. If you plan to continue using the project after the walkthrough, you may keep these configurations in place.
|
||||
|
||||
However, if you decide to remove them after completing the walkthrough, you can use the following gcloud commands:
|
||||
> <walkthrough-pin-section-icon></walkthrough-pin-section-icon> These commands will delete the firewall rules and network configurations created during the walkthrough. Make sure to run them only if you no longer need these settings.
|
||||
|
||||
```
|
||||
gcloud compute firewall-rules delete allow-tcp-8080 --quiet
|
||||
|
||||
gcloud compute firewall-rules delete allow-iap --quiet
|
||||
|
||||
gcloud compute networks subnets delete default --region us-central1 --quiet
|
||||
|
||||
gcloud compute networks delete default --quiet
|
||||
```
|
||||
Loading…
Add table
Add a link
Reference in a new issue