-
Notifications
You must be signed in to change notification settings - Fork 35
Cost estimate
Abhijeet Bodhankar edited this page Feb 4, 2020
·
3 revisions
The estimate below assumes:
- 500 users in the tenant.
- 100 buildings and 100 rooms per building.
The recommended SKU for a production environment is:
- App Service: Standard (S1)
- Azure Search: Basic - The Azure Search service cannot be upgraded once it is provisioned, so select a tier that will meet your anticipated needs.
Data storage: 1 GB max.
Table data operations:
-
Azure table storage 'UserFavorites' is used to store user favorites room data
- Total number of read calls for UserFavorites table = 500 users * 200 calls to fetch favorite rooms /user/month = 10000 favorite rooms fetch calls/month.
- Total number of write calls for UserFavorites table = 500 users * 5 calls to update favorite rooms /user/month = 2500 calls to update favorite rooms/month.
-
Azure table storage 'RoomCollection' is used to store rooms and buildings information synced from Exchange using Azure Function.
- Total number of read calls for RoomCollection table = 1 Function app * 4 weeks * 100 call to fetch rooms of 100 buildings/week = 400 calls to fetch all rooms/month.
- Total number of delete calls for RoomCollection table = 1 Function app * 4 weeks * 100 call to delete rooms of 100 buildings/week = 400 calls to fetch all rooms/month.
- Total number of write calls for RoomCollection table = 1 Function app * 4 weeks * 100 call to insert rooms of 100 buildings/week = 400 calls to update all rooms/month.
- Minimal calls required to remove rooms from user favorites as possibility of room deletion from Microsoft Exchange frequently is less.
-
Azure table storage 'UserConfiguration' is used to store user local time zone.
- Total number of read calls for UserConfiguration table = 500 users * 200 calls to fetch user local time zone /user/month = 10000 fetch calls/month.
- Total number of write calls for UserConfiguration table = 500 users * 2 calls to fetch user local time zone /user/month = 1000 fetch calls/month.
Blob data operations:
- Blob stores the bot state to check whether welcome card is sent to the user. Blob storage has container name as "bot-state".
- Total number of read calls in storage for welcome card validation = 500 users * 1 welcome message = 500 operations (occurs just once during application installation)
- Total number of write calls in storage for welcome card validation = 500 users * 1 welcome message = 500 operations (occurs just once during application installation)
- Total number of write calls in storage to store conversation state by bot = 500 users * 20 operations /user/day = 10000 calls per month
IMPORTANT: This is only an estimate, based on the assumptions above. Your actual costs may vary.
Prices were taken from the Azure Pricing Overview on 01 Jan 2020, for the West US 2 region.
Use the Azure Pricing Calculator to model different service tiers and usage patterns.
Resource | Tier | Load | Monthly price |
---|---|---|---|
Storage Account (Table) | Standard_LRS | < 1GB data, 24,700 operations (Cost of table storage account will be same till 75,000 operations) | $0.06 |
Storage account (Blob) | Standard_LRS | < 1GB data, 10,500 write operations, 500 read operations | $0.06 |
Bot Channels Registration | F0 | N/A | Free |
App Service Plan | S2 | 744 hours | $148.80 |
App Service (Bot) | - | (charged to App Service Plan) | |
Azure Search | B | 744 hours | $75.14 |
Application Insights (Bot) | (free up to 5 GB) | ||
Azure Function | Dedicated | 100 Action Executions for 31 days | (free up to 1 million executions) |
Total | $224 |