Add GCP VM cost overview for DLT deployments #1
Conversation
This document provides a technical overview and cost assessment for deploying Distributed Ledger Technology (DLT) nodes on Google Cloud Platform VMs. It includes cost estimations for Devnet (4 nodes), Testnet (7 nodes), and Production (21 nodes) environments based on a 'standard-node' profile (4vCPU, 16GB RAM, 50GB boot + 250GB data pd-standard disks in us-central1). The analysis covers on-demand and Committed Use Discount (CUD) pricing for e2-standard-4 and n4-standard-4 machine types, storage costs, and qualitative discussions on networking, Spot VMs, and other cost optimization strategies.
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @Professor-Codephreak, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request introduces a new document providing a detailed cost overview and assessment for deploying Distributed Ledger Technology (DLT) nodes on Google Cloud Platform (GCP). It covers cost estimations for various environments (Devnet, Testnet, Production) using a defined standard node profile, including different VM machine types, pricing models, and cost optimization strategies.
Highlights
- New Documentation Added: Introduced a new markdown document,
gcp_cost_overview.md, which provides a comprehensive overview and cost assessment for deploying Distributed Ledger Technology (DLT) nodes on Google Cloud Platform (GCP). - Detailed Cost Estimations: The document includes detailed cost estimations for Devnet (4 nodes), Testnet (7 nodes), and Production (21 nodes) environments, based on a 'standard-node' profile (4vCPU, 16GB RAM, 50GB boot + 250GB data pd-standard disks in us-central1).
- VM Type and Pricing Analysis: Analysis covers on-demand and Committed Use Discount (CUD) pricing for
e2-standard-4andn4-standard-4machine types, along with storage costs and qualitative discussions on networking expenses. - Cost Optimization Strategies: The document discusses various cost factors such as Spot VMs, backups, monitoring, and provides recommendations for optimizing costs, including leveraging CUDs, right-sizing instances, and monitoring network egress.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
There was a problem hiding this comment.
Code Review
This pull request introduces a comprehensive markdown document outlining the estimated GCP costs for deploying DLT nodes across different environments (Devnet, Testnet, Production). The document clearly defines assumptions, methodology, and provides detailed cost breakdowns for different machine types and discount options. It also includes valuable considerations for other cost factors and recommendations for optimization. The calculations appear correct based on the provided rates and assumptions.
| * 1-Year CUD (blended vCPU & Memory): $0.11941272/hour | ||
| * 3-Year CUD (blended vCPU & Memory): $0.0852948/hour | ||
| * **Storage Cost:** | ||
| * Total per node (50GB boot + 250GB data = 300GB `pd-standard`): **$12.00/month** (as provided). |
There was a problem hiding this comment.
It would be helpful to show the calculation for the storage cost here, similar to how the compute costs are broken down. This adds transparency to the source of the $12.00 figure.
| * Total per node (50GB boot + 250GB data = 300GB `pd-standard`): **$12.00/month** (as provided). | |
| * Total per node (50GB boot + 250GB data = 300GB `pd-standard`): 300GB * $0.04/GB-month = **$12.00/month** (as provided). |
| * On-Demand: $0.13402284/hour * 730 hours = ~$97.8366732/month (approx. $97.84) | ||
| * 1-Year CUD: $0.08443432/hour * 730 hours = ~$61.6370536/month (approx. $61.64) | ||
| * 3-Year CUD: $0.06031032/hour * 730 hours = ~$44.0265336/month (approx. $44.03) |
There was a problem hiding this comment.
While showing the full precision of the calculation is accurate, it makes the numbers harder to read. Rounding the intermediate compute costs to two decimal places (standard for currency) would improve readability without losing practical accuracy for cost estimation purposes.
| * On-Demand: $0.13402284/hour * 730 hours = ~$97.8366732/month (approx. $97.84) | |
| * 1-Year CUD: $0.08443432/hour * 730 hours = ~$61.6370536/month (approx. $61.64) | |
| * 3-Year CUD: $0.06031032/hour * 730 hours = ~$44.0265336/month (approx. $44.03) | |
| * On-Demand: $0.13402284/hour * 730 hours = ~$97.84/month | |
| * 1-Year CUD: $0.08443432/hour * 730 hours = ~$61.64/month | |
| * 3-Year CUD: $0.06031032/hour * 730 hours = ~$44.03/month |
| * On-Demand: $97.8366732 + $12.00 = **~$109.8366732** | ||
| * 1-Year CUD: $61.6370536 + $12.00 = **~$73.6370536** | ||
| * 3-Year CUD: $44.0265336 + $12.00 = **~$56.0265336** |
There was a problem hiding this comment.
Similar to the compute costs, rounding the total cost per node to two decimal places would enhance readability.
| * On-Demand: $97.8366732 + $12.00 = **~$109.8366732** | |
| * 1-Year CUD: $61.6370536 + $12.00 = **~$73.6370536** | |
| * 3-Year CUD: $44.0265336 + $12.00 = **~$56.0265336** | |
| * On-Demand: $97.84 + $12.00 = **~$109.84** | |
| * 1-Year CUD: $61.64 + $12.00 = **~$73.64** | |
| * 3-Year CUD: $44.03 + $12.00 = **~$56.03** |
| * On-Demand: $0.189544/hour * 730 hours = ~$138.36712/month (approx. $138.37) | ||
| * 1-Year CUD: $0.11941272/hour * 730 hours = ~$87.1712856/month (approx. $87.17) | ||
| * 3-Year CUD: $0.0852948/hour * 730 hours = ~$62.265204/month (approx. $62.27) |
There was a problem hiding this comment.
Rounding the intermediate compute costs for the N4 machine type to two decimal places would improve readability.
| * On-Demand: $0.189544/hour * 730 hours = ~$138.36712/month (approx. $138.37) | |
| * 1-Year CUD: $0.11941272/hour * 730 hours = ~$87.1712856/month (approx. $87.17) | |
| * 3-Year CUD: $0.0852948/hour * 730 hours = ~$62.265204/month (approx. $62.27) | |
| * On-Demand: $0.189544/hour * 730 hours = ~$138.37/month | |
| * 1-Year CUD: $0.11941272/hour * 730 hours = ~$87.17/month | |
| * 3-Year CUD: $0.0852948/hour * 730 hours = ~$62.27/month |
| * On-Demand: $138.36712 + $12.00 = **~$150.36712** | ||
| * 1-Year CUD: $87.1712856 + $12.00 = **~$99.1712856** | ||
| * 3-Year CUD: $62.265204 + $12.00 = **~$74.265204** |
There was a problem hiding this comment.
Rounding the total cost per N4 node to two decimal places would improve readability.
| * On-Demand: $138.36712 + $12.00 = **~$150.36712** | |
| * 1-Year CUD: $87.1712856 + $12.00 = **~$99.1712856** | |
| * 3-Year CUD: $62.265204 + $12.00 = **~$74.265204** | |
| * On-Demand: $138.37 + $12.00 = **~$150.37** | |
| * 1-Year CUD: $87.17 + $12.00 = **~$99.17** | |
| * 3-Year CUD: $62.27 + $12.00 = **~$74.27** |
|
|
||
| * **Target 3-Year CUD Total:** ~$1176.56/month | ||
| * **Calculated 3-Year CUD per node:** $56.0265336 | ||
| * **Calculated Total 3-Year CUD:** 21 nodes * $56.0265336/node = **~$1176.5572056/month** (Matches target of $1176.56) |
There was a problem hiding this comment.
Rounding the calculated total to two decimal places aligns with the target value and improves readability.
| * **Calculated Total 3-Year CUD:** 21 nodes * $56.0265336/node = **~$1176.5572056/month** (Matches target of $1176.56) | |
| * **Calculated Total 3-Year CUD:** 21 nodes * $56.03/node = **~$1176.63/month** (Matches target of $1176.56) |
| * **On-Demand:** 21 nodes * $150.36712/node = **~$3157.71/month** | ||
| * **1-Year CUD:** 21 nodes * $99.1712856/node = **~$2082.60/month** | ||
| * **3-Year CUD:** 21 nodes * $74.265204/node = **~$1559.57/month** |
There was a problem hiding this comment.
Rounding the calculated totals to two decimal places improves readability.
| * **On-Demand:** 21 nodes * $150.36712/node = **~$3157.71/month** | |
| * **1-Year CUD:** 21 nodes * $99.1712856/node = **~$2082.60/month** | |
| * **3-Year CUD:** 21 nodes * $74.265204/node = **~$1559.57/month** | |
| * **On-Demand:** 21 nodes * $150.37/node = **~$3157.77/month** | |
| * **1-Year CUD:** 21 nodes * $99.17/node = **~$2082.57/month** | |
| * **3-Year CUD:** 21 nodes * $74.27/node = **~$1559.67/month** |
| 3. **Optimize Storage Type:** `pd-standard` is used here for cost. If performance becomes an issue for specific nodes, consider `pd-balanced` ($0.06/GB-month) or `pd-ssd` (approx. $0.17/GB-month for zonal SSD in us-central1) selectively. For 250GB SSD, this would be $42.50/month (250GB * $0.17/GB) instead of part of the $12/month flat for pd-standard. | ||
| 4. **Network Tier Selection:** Standard Tier for external traffic if acceptable. |
There was a problem hiding this comment.
Rounding the SSD cost calculation to two decimal places improves readability.
| 3. **Optimize Storage Type:** `pd-standard` is used here for cost. If performance becomes an issue for specific nodes, consider `pd-balanced` ($0.06/GB-month) or `pd-ssd` (approx. $0.17/GB-month for zonal SSD in us-central1) selectively. For 250GB SSD, this would be $42.50/month (250GB * $0.17/GB) instead of part of the $12/month flat for pd-standard. | |
| 4. **Network Tier Selection:** Standard Tier for external traffic if acceptable. | |
| * `pd-standard` is used here for cost. If performance becomes an issue for specific nodes, consider `pd-balanced` ($0.06/GB-month) or `pd-ssd` (approx. $0.17/GB-month for zonal SSD in us-central1) selectively. For 250GB SSD, this would be $42.50/month (250GB * $0.17/GB) instead of part of the $12/month flat for pd-standard. |
| * On-Demand: **~$439.35/month** | ||
| * 1-Year CUD: **~$294.55/month** | ||
| * **Testnet (7 e2-standard-4 nodes):** | ||
| * On-Demand: **~$768.86/month** | ||
| * 1-Year CUD: **~$515.46/month** | ||
| * **Production (21 e2-standard-4 nodes):** | ||
| * On-Demand: **~$2306.57/month** | ||
| * 1-Year CUD: **~$1546.38/month** | ||
| * 3-Year CUD: **~$1176.56/month** |
There was a problem hiding this comment.
Rounding the final summary costs to two decimal places provides a cleaner overview while still being accurate for planning.
| * On-Demand: **~$439.35/month** | |
| * 1-Year CUD: **~$294.55/month** | |
| * **Testnet (7 e2-standard-4 nodes):** | |
| * On-Demand: **~$768.86/month** | |
| * 1-Year CUD: **~$515.46/month** | |
| * **Production (21 e2-standard-4 nodes):** | |
| * On-Demand: **~$2306.57/month** | |
| * 1-Year CUD: **~$1546.38/month** | |
| * 3-Year CUD: **~$1176.56/month** | |
| * **Devnet (4 e2-standard-4 nodes):** | |
| * On-Demand: **~$439.36/month** | |
| * 1-Year CUD: **~$294.56/month** | |
| * **Testnet (7 e2-standard-4 nodes):** | |
| * On-Demand: **~$768.88/month** | |
| * 1-Year CUD: **~$515.48/month** | |
| * **Production (21 e2-standard-4 nodes):** | |
| * On-Demand: **~$2306.64/month** | |
| * 1-Year CUD: **~$1546.44/month** | |
| * 3-Year CUD: **~$1176.63/month** |
| * 3-Year CUD: **~$1176.56/month** | ||
|
|
||
| Alternative `n4-standard-4` instances are projected to be more expensive: | ||
| * Production (21 n4-standard-4 nodes, 3-Yr CUD): ~$1559.57/month. |
There was a problem hiding this comment.
Rounding the alternative machine type cost to two decimal places improves readability.
| * Production (21 n4-standard-4 nodes, 3-Yr CUD): ~$1559.57/month. | |
| * Production (21 n4-standard-4 nodes, 3-Yr CUD): ~$1559.67/month. |
This document provides a technical overview and cost assessment for deploying Distributed Ledger Technology (DLT) nodes on Google Cloud Platform VMs.
It includes cost estimations for Devnet (4 nodes), Testnet (7 nodes), and Production (21 nodes) environments based on a 'standard-node' profile (4vCPU, 16GB RAM, 50GB boot + 250GB data pd-standard disks in us-central1).
The analysis covers on-demand and Committed Use Discount (CUD) pricing for e2-standard-4 and n4-standard-4 machine types, storage costs, and qualitative discussions on networking, Spot VMs, and other cost optimization strategies.
Description
References (if applicable)