Satte på en film uten å sjekke lmdb score først, living on the edge
August 27, 2026 at 3:17 PM
Satte på en film uten å sjekke lmdb score først, living on the edge
You need LMDB aka @lisamarietea.bsky.social
December 31, 2024 at 11:08 PM
You need LMDB aka @lisamarietea.bsky.social
4. Analytics (DuckDB, ClickHouse, DataFusion CLI, Databend)
5. Time series (Timescale, QuestDB, Influx)
6. Sharded, distributed key-value (FoundationDB, TiKV)
7. Non-sharded, distributed key-value (etcd, rqlite, Tigerbeetle, zookeeper)
8. Embedded key-value (RocksDB, BerkeleyDB, LMDB)
5. Time series (Timescale, QuestDB, Influx)
6. Sharded, distributed key-value (FoundationDB, TiKV)
7. Non-sharded, distributed key-value (etcd, rqlite, Tigerbeetle, zookeeper)
8. Embedded key-value (RocksDB, BerkeleyDB, LMDB)
December 1, 2024 at 3:28 PM
4. Analytics (DuckDB, ClickHouse, DataFusion CLI, Databend)
5. Time series (Timescale, QuestDB, Influx)
6. Sharded, distributed key-value (FoundationDB, TiKV)
7. Non-sharded, distributed key-value (etcd, rqlite, Tigerbeetle, zookeeper)
8. Embedded key-value (RocksDB, BerkeleyDB, LMDB)
5. Time series (Timescale, QuestDB, Influx)
6. Sharded, distributed key-value (FoundationDB, TiKV)
7. Non-sharded, distributed key-value (etcd, rqlite, Tigerbeetle, zookeeper)
8. Embedded key-value (RocksDB, BerkeleyDB, LMDB)
Hey 👋 It’s been a while! This morning, we published a new article about how we made @meilisearch.com’s semantic search much faster with hannoy. Hannoy is a new LMDB disk-based HNSW vector store that is much more perf. Now, it’s the default backend! #rustlang blog.kerollmops.com/from-trees-t...
From trees to graphs: speeding up vector search 10x with Hannoy
This is a walkthrough of a new LMDB-based HNSW Rust vector store with drastically better performances than the previous random-projection-based one. This is a disk-based HNSW highly vector store inspi...
blog.kerollmops.com
December 8, 2025 at 4:07 PM
Hey 👋 It’s been a while! This morning, we published a new article about how we made @meilisearch.com’s semantic search much faster with hannoy. Hannoy is a new LMDB disk-based HNSW vector store that is much more perf. Now, it’s the default backend! #rustlang blog.kerollmops.com/from-trees-t...
lmdb is cool!
I tried making the LLLL (linux, lwan, lmdb, lua) stack to stick, but people don't seem to like choosing their stack based solely on an alliteration
I tried making the LLLL (linux, lwan, lmdb, lua) stack to stick, but people don't seem to like choosing their stack based solely on an alliteration
October 26, 2023 at 4:58 PM
lmdb is cool!
I tried making the LLLL (linux, lwan, lmdb, lua) stack to stick, but people don't seem to like choosing their stack based solely on an alliteration
I tried making the LLLL (linux, lwan, lmdb, lua) stack to stick, but people don't seem to like choosing their stack based solely on an alliteration
what's the best kind of storage for an mst? i was thinking something like heed/lmdb or rsdb
July 27, 2025 at 12:40 AM
what's the best kind of storage for an mst? i was thinking something like heed/lmdb or rsdb
I'm like LMDB but for political scandals involving burlesque dancers
May 30, 2024 at 9:44 PM
I'm like LMDB but for political scandals involving burlesque dancers
@anthropologal.bsky.social calls me LMdB bc if it's Hollywood related, I'm gonna pull it from the depths of my dumb brain.
August 19, 2025 at 5:11 PM
@anthropologal.bsky.social calls me LMdB bc if it's Hollywood related, I'm gonna pull it from the depths of my dumb brain.
More people should ditch redis and just use lmdb 👌
December 15, 2025 at 12:22 PM
More people should ditch redis and just use lmdb 👌
Have you ran these benchmarks for heed/LMDB too? 👀
October 2, 2025 at 5:38 PM
Have you ran these benchmarks for heed/LMDB too? 👀
Ho crap! That's unfortunate, because LMDB is often in the top performers
October 2, 2025 at 5:42 PM
Ho crap! That's unfortunate, because LMDB is often in the top performers
my therapist said i should put the blobs in lmdb and when i said that will make keeping it in sync difficult because the queue is in postgres they ended the session early
December 28, 2025 at 9:21 AM
my therapist said i should put the blobs in lmdb and when i said that will make keeping it in sync difficult because the queue is in postgres they ended the session early
Hoping to share more soon, but last week I developed a workflow that handles complex relationship extraction from any Wikipedia page of any size. I'll be making this open source soon. We did a few tricks to make it really fast (hint, we rebuilt Wikipedia and Wikidata both locally with LMDB).
August 4, 2025 at 10:13 PM
Hoping to share more soon, but last week I developed a workflow that handles complex relationship extraction from any Wikipedia page of any size. I'll be making this open source soon. We did a few tricks to make it really fast (hint, we rebuilt Wikipedia and Wikidata both locally with LMDB).
languages like that just make me sad though because they don't really deal practical without community support
i had to migrate from lmdb to sqlite :(
i had to migrate from lmdb to sqlite :(
July 29, 2026 at 3:30 AM
languages like that just make me sad though because they don't really deal practical without community support
i had to migrate from lmdb to sqlite :(
i had to migrate from lmdb to sqlite :(
Bf* Tree
* f stands for faster
This made me chuckle for some reason, but the paper is interesting. I'd be interested to see how it compares to LMDB performance (I know not a 1to1 comparison per se)
* f stands for faster
This made me chuckle for some reason, but the paper is interesting. I'd be interested to see how it compares to LMDB performance (I know not a 1to1 comparison per se)
Bf-trees are a high-performance alternative to the classic B-tree.
The key contribution: Decoupling the size of on-disk pages from the size of memory cache entries.
I'm curious if/when MySQL and Postgres will adopt modern versions of classic indexes.
The key contribution: Decoupling the size of on-disk pages from the size of memory cache entries.
I'm curious if/when MySQL and Postgres will adopt modern versions of classic indexes.
December 31, 2025 at 1:58 AM
Bf* Tree
* f stands for faster
This made me chuckle for some reason, but the paper is interesting. I'd be interested to see how it compares to LMDB performance (I know not a 1to1 comparison per se)
* f stands for faster
This made me chuckle for some reason, but the paper is interesting. I'd be interested to see how it compares to LMDB performance (I know not a 1to1 comparison per se)
And if you’re interested, check out the repo!
I tried to document internal algorithms github.com/meilisearch/...
I tried to document internal algorithms github.com/meilisearch/...
October 14, 2025 at 8:19 PM
And if you’re interested, check out the repo!
I tried to document internal algorithms github.com/meilisearch/...
I tried to document internal algorithms github.com/meilisearch/...
'Just saying... LMDB aborts the process in case an assertion is invalid. But it can be disabled if you compile with NDEBUG which will make LMDB continue to run in an invalid state... fosstodon.org/@kero/114137...
March 10, 2025 at 9:19 AM
'Just saying... LMDB aborts the process in case an assertion is invalid. But it can be disabled if you compile with NDEBUG which will make LMDB continue to run in an invalid state... fosstodon.org/@kero/114137...
hey @krisajenkins.bsky.social, you should have Hauhai Yang on your show to talk about the 1.0 release of his datalog database Datalevin. he wrote a whole book on the subject, i suspect he has a lot to say about it, and i know you'd ask great questions.
github.com/huahaiy
datalevin.org
github.com/huahaiy
datalevin.org
Datalevin Docs
Datalevin documentation, examples, and guides for Datalog, LMDB-backed storage, search, vectors, and production deployment.
datalevin.org
September 15, 2026 at 10:21 AM
hey @krisajenkins.bsky.social, you should have Hauhai Yang on your show to talk about the 1.0 release of his datalog database Datalevin. he wrote a whole book on the subject, i suspect he has a lot to say about it, and i know you'd ask great questions.
github.com/huahaiy
datalevin.org
github.com/huahaiy
datalevin.org
Ahora que somos pocos, os confieso que una vez, buscando una peli que fuera buena, abrimos un vinito y empezamos a leer sinopsis, muy atentos a la valoración que le daban los de lMDb.
Media tarde y botella después, nos percatamos de que lo que estábamos mirando era esto 👇
Media tarde y botella después, nos percatamos de que lo que estábamos mirando era esto 👇
September 20, 2023 at 10:01 PM
Ahora que somos pocos, os confieso que una vez, buscando una peli que fuera buena, abrimos un vinito y empezamos a leer sinopsis, muy atentos a la valoración que le daban los de lMDb.
Media tarde y botella después, nos percatamos de que lo que estábamos mirando era esto 👇
Media tarde y botella después, nos percatamos de que lo que estábamos mirando era esto 👇
the current status is that these bindings (for ruby, for a little direct-attached key-value database called lmdb, the storage engine that ships with openldap) are actually in a lot better shape than they were yesterday morning
i will be cutting a new release shortly…
github.com/doriantaylor...
i will be cutting a new release shortly…
github.com/doriantaylor...
github.com
June 17, 2026 at 1:22 PM
the current status is that these bindings (for ruby, for a little direct-attached key-value database called lmdb, the storage engine that ships with openldap) are actually in a lot better shape than they were yesterday morning
i will be cutting a new release shortly…
github.com/doriantaylor...
i will be cutting a new release shortly…
github.com/doriantaylor...
Without having numbers (right now), LMDB will probably read absurdly fast, while not so much space- or write efficient.
Most, if not all data, fits into RAM here. Would be interesting to benchmark 100s of GBs.
But I found redb to be slower (sometimes much slower) than LMDB, pretty much always.
Most, if not all data, fits into RAM here. Would be interesting to benchmark 100s of GBs.
But I found redb to be slower (sometimes much slower) than LMDB, pretty much always.
October 11, 2025 at 2:22 PM
Without having numbers (right now), LMDB will probably read absurdly fast, while not so much space- or write efficient.
Most, if not all data, fits into RAM here. Would be interesting to benchmark 100s of GBs.
But I found redb to be slower (sometimes much slower) than LMDB, pretty much always.
Most, if not all data, fits into RAM here. Would be interesting to benchmark 100s of GBs.
But I found redb to be slower (sometimes much slower) than LMDB, pretty much always.