Skip to content

Add replication-latency description#39

Open
kriszyp wants to merge 2 commits intomainfrom
replication-latency-description
Open

Add replication-latency description#39
kriszyp wants to merge 2 commits intomainfrom
replication-latency-description

Conversation

@kriszyp
Copy link
Copy Markdown
Member

@kriszyp kriszyp commented Dec 5, 2025

No description provided.

@kriszyp kriszyp requested a review from cap10morgan December 5, 2025 15:39
Copy link
Copy Markdown
Contributor

@cap10morgan cap10morgan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Couple little issues to address, but thanks for adding this!

Comment thread README.md Outdated
| `cluster_disconnected_dbs` | Count of disconnected dbs by node. |
| `replication_backlog` | Number of pending replication consumers by origin, database, & table (for NATS replication only). |
| `replication_backlog_time` | The difference in milliseconds between lastReceivedRemoteTime and lastReceivedLocalTime for each of the node's dbs (for Harper replication only). |
| `replication_backlog_time` | The difference in milliseconds between lastReceivedRemoteTime and lastReceivedLocalTime for each of the node's dbs (for Harper replication only). | | `replication_latency` | The difference in milliseconds between incoming transactions' timestamps (as record remotely) and the time when each transaction finished committing locally (for Harper replication only). |
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks like there's an | at the beginning and should as record remotely be as recorded remotely?

@hdbjeff
Copy link
Copy Markdown

hdbjeff commented Dec 5, 2025

Could we also add what is different about this vs replication_backlog_time? It sounds like this also includes the time to commit the received transaction, whereas replication_backlog_time does not?

@kriszyp
Copy link
Copy Markdown
Member Author

kriszyp commented Dec 8, 2025

replication_latency is an actual Harper analytic recorded over time. I think replication_backlog_time is prometheus guage computed from a point in time reading of cluster_status. Does that sound right Wes?

@cap10morgan
Copy link
Copy Markdown
Contributor

replication_latency is an actual Harper analytic recorded over time. I think replication_backlog_time is prometheus guage computed from a point in time reading of cluster_status. Does that sound right Wes?

Yes. replication_backlog_time is computed from the response_time attribute of the cluster_network op whereas replication_latency comes from an hdb_analytics query.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants