neon icon indicating copy to clipboard operation
neon copied to clipboard

Neon: Serverless Postgres. We separated storage and compute to offer autoscaling, code-like database branching, and scale to zero.

Results 520 neon issues
Sort by recently updated
recently updated
newest added

The "manual_gc" context was removed in commit be0c73f8e7. The code that generated the other error was removed in commit 9a6c0be823.

Per the "Cleaning up parent-shard layers" section in #6358 -- currently after a shard split, layers from the parent shards are not deleted until the whole tenant is eventually deleted....

c/storage/pageserver
a/tech_debt

## Problem NodeInfo is 256 bytes. Old EndpointCacheKey is more expensive to hash and compare given it's string based. ## Summary of changes NodeCachedInfo is 48 bytes. New EndpointCacheKey is...

Currently, when we do an `InMemoryLayer::write_to_disk`, there is a tremendous amount of random read I/O, as deltas from the ephemeral file (written in LSN order) are written out to the...

c/storage/pageserver
a/tech_debt

Add support for event triggers to capture DDL events: https://www.postgresql.org/docs/current/event-triggers.html We currently do not support event triggers due to the superuser requirement. Related slack thread: https://neondb.slack.com/archives/C03QLRH7PPD/p1710592909306899 User requests: https://discord.com/channels/1176467419317940276/1216517379366846474/1216517382898454639 https://community.neon.tech/t/permission-denied-to-create-event-trigger/443

t/feature
c/PostgreSQL
c/compute

The `compact_level0_phase1` currently uses `ValueRef::load` here, which internally uses `read_blob` with the `FileBlockReader` against the delta layer's VirtualFiles. This still goes through the `PageCache` for the data pages. (We do...

part of epic https://github.com/neondatabase/neon/issues/7386 bit of prior discussion in https://neondb.slack.com/archives/C033RQ5SPDH/p1719411245662839 --- `InMemoryLayer::get_values_reconstruct_data` uses `read_blob`, which internally uses the PageCache for block access. Switch it to vectored reads that bypass the...

We should check it always after a sleep even before starting streaming.

## Problem vm-monitor limits the size of the LFC not go beyond `total_memory - reserved_memory`. With the not-that-recent experiment result, we can give customers better performance by increasing the LFC...

## Problem Scale test doesn't exercise drain & fill. ## Summary of changes Make scale test exercise drain & fill ## Checklist before requesting a review - [ ] I...

run-benchmarks