Data Tiers prototypes

Index lifecycle policies

Let's elevate the prominence of the default policies to make it easier for users to follow best practices. Let's hide the custom policies so that advanced users can still access them, but they don't compete with the default policies that address 80% of use cases.

Full spectrum

Details
Tiers
Hot
Warm
Cold
Frozen
Used by
5

Hot-warm

Details
Tiers
Hot
Warm
Cold
Frozen
Used by
38

Deep-freeze

Details
Tiers
Hot
Warm
Cold
Frozen
Used by
11

Create index lifecycle policy

How do we emphasize the sequential nature of data tiers in the UI? How do we help the user differentiate between the steps? How do we give them an overview of how this policy behaves?

Timing comparison

Hot tier
Active
Rollover
Enabled
Maximum index size
50gb
Maximum age
30 days
Index priority
1
Actions
Shrink
Force merge
Warm tier
Active
Rollover
Enabled
Maximum index size
50gb
Maximum age
30 days
Index priority
1
Actions
Shrink
Force merge
Cold tier
Active

Cold tier not active

Frozen tier
Active
Rollover
Enabled
Maximum index size
50gb
Maximum age
30 days
Index priority
1
Actions
Shrink
Force merge

Data streams

What do data tiers look like in the context of data streams? What would users find useful about this?

Name
Indices
Errors
Name
data-stream-1
Indices
125
Errors
19
Name
data-stream-2
Indices
2922
Errors
0
Name
data-stream-3
Indices
56
Errors
3
Name
data-stream-4
Indices
0
Errors
0
Name
data-stream-5
Indices
438
Errors
0

Cluster supervisor

What does the cluster state look like through the lens of data tiers? What would users find useful about this?

Hot tier

Configure
Performance
Fast
Cost
Highest
Duration
Weeks
Nodes
4
Indices
12182

Warm tier

Configure
Performance
Medium to fast
Cost
Moderate
Duration
Weeks
Nodes
2
Indices
34552

Cold tier

Inactive
Activate
Performance
Slow to fast
Cost
Low
Duration
Months

Frozen tier

Configure
Performance
Very slow
Cost
Lowest
Duration
Years
Nodes
1
Indices
786323