Update HA page: unresponsive endpoint detection and node failure fixes #2878
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
bgrenon
changed the title
danieltprice
reviewed
Feb 4, 2025
danieltprice
reviewed
Feb 4, 2025
danieltprice
reviewed
Feb 4, 2025
danieltprice
reviewed
Feb 4, 2025
|
|
||
| If a compute endpoint is in a degraded state (repeatedly crashing and restarting rather than failing outright), we will detect and reattach it automatically, typically within 5 minutes. During this time, your application may experience intermittent connectivity. | ||
|
|
||
| #### Node failures |
There was a problem hiding this comment.
same comment as above.
danieltprice
reviewed
Feb 4, 2025
danieltprice
reviewed
Feb 4, 2025
| | VM failure | Brief interruption | VM recreation and endpoint reattachment | Seconds | | ||
| | Degraded endpoint | Possible intermittent connectivity | Automatic detection and reattachment | Up to 5 minutes | | ||
| | Node failure | Compute unavailable | Rescheduling to healthy nodes | ~2 minutes | | ||
|
|
||
| ### Impact on session data after a failure? | ||
|
|
||
| While your application should handle reconnections automatically, session-specific data like temporary tables, prepared statements, and the Local File Cache ([LFC](/docs/reference/glossary#local-file-cache)), which stores frequently accessed data, will not persist across a failover. As a result, queries may initially run more slowly until the Postgres memory buffers and cache are rebuilt. |
There was a problem hiding this comment.
We use "failover" above. Should we use "recovery"? For the sake of argument, why can't we call this section "Compute failover" as some have suggested and explain that failover in Neon's serverless architecture is a little different from traditional failover.
need confirmation of regions that support multi-AZ
…ebsite into bgrenon-ha-update
bgrenon
changed the title
Co-authored-by: Daniel <10074684+danieltprice@users.noreply.github.com>
…ebsite into bgrenon-ha-update
danieltprice
approved these changes
Feb 12, 2025
…ebsite into bgrenon-ha-update
bgrenon
requested review from
ololobus,
mickael-carl,
sharnoff and
vadim2404
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This page was updated mostly to add "compute failover' sections for:
Preview: https://neon-next-git-bgrenon-ha-update-neondatabase.vercel.app/docs/introduction/high-availability#compute-failover