<aside> ➡️ Building a Real Time Metrics Database at Datadog

</aside>

https://leetcode.com/discuss/interview-question/system-design/287678/design-a-monitoring-or-analytics-service-like-datadog-or-signalfx

1. Problem statement

Use cases

• User submits count and guage metrics to system
• User creates monitors and alerts
• System send triggered monitors and alerts to users
• User views metrics and graphs in UI
• User can tag metrics

Out of scope

• Public APIs for querying

Constraints and assumptions

• Per-customer ballpark numbers
  • $10^4$ apps (1,000s of hosts * 10s containers)
  • $10^3$ metrics emitted from each app/container
  • $10^0$ (1) point a second per metric
  • $10^5$ secs per day (actually 86,400)
  • $10^1$ bytes/point (8 byte float, amortized tags)
  • $10^4 * 10^3 * 10^0 * 10^5 * 10^1 = 10^{13}$
  • 10 TB a day for one customer
• Number of customers = $10^4 = 10,000$
• $10^{13} * 10^4 = 10^{17} = 100PB$ a day
• 1000 queries per customer per day

2. Clarifying questions

todo

3. Functional requirements

• User
  • Write stat → POST /api/stats:write
  • Describe tags
    • What tags are queryable for this metric?
  • Tag index
    • Give a time series ID, what tags were used?
  • Tag inverted index
    • Given some tags and a time range, what were the time series ingested?
  • Point store
    • What are the values of a time series between two times?