pragmaxim
pragmaxim
``` case class BlockCache(byId: Map[BlockId, BlockInfo], byHeight: SortedMap[Int, BlockInfo]) case class ProgressState( lastBlockIdInEpoch: SortedMap[Int, BlockId], invalidEpochs: SortedMap[Int, InvalidEpochCandidate], blockCache: BlockCache ) ``` ``` val x = stateRef.transformAndExtract { state =>...
Hey, I've done some profiling as indexing data significantly slowed down after I added `upsert` operation in postgresql mode, and turns out that `throw getDuplicateKeyException(k.toString());` in `MVSecondaryIndex#checkUnique` is extremely expensive...
Hey, I'm building a multi-map storage that needs a lot of secondary indexes so in : ``` xByY: MVMap, yByZ: MVMap ``` I'd like to replace `Y` for its index...
Hi, This might improve performance a lot in both query time and indexing time. Problem : - ErgoTreeTemplate is created here for ALL ErgoTrees instead of just P2SH. The Template...
Hi, just out of curiosity, why you haven't implemented adding new elements to an existing instance? As it does not seem to be too challenging or am I missing something?
Scylla version 5.0.1-0.20220719.b177dacd3 nodetool ReleaseVersion: 3.0.8 From https://hub.docker.com/r/scylladb/scylla/ Having : ``` CREATE KEYSPACE keyspaceFoo WITH replication = {'class': 'SimpleStrategy', 'replication_factor': '1'}; ``` And : ``` nodetool disableautocompaction keyspaceFoo tableX tableY...
`org.reactivestreams.Processor` is an equivalent of `ZPipeline` or `akka.stream.scaladsl.Flow` and when integrating between different reactive "implementation", it is possible to create it using `org.reactivestreams.FlowAdapters.toFlowProcessor` ... Akka-stream has `flow.toProcessor` method but I...
Regarding this comment https://github.com/ergoplatform/ergo/pull/1549#discussion_r777116646 We could have functions accepting `InputLike` when they are only interested in `boxId`
### Expected behavior `db.flush` should flush memtables to sst files. I use it in a shutdown hook so that I don't loose data as I am disabling WAL for performance...
Repro https://github.com/pragmaxim-com/rust-playground/commit/e5ec1c6956d8868a6e09257fc71fb40fada2bab4 `db.flush` call does not lead to new (the first) sst file being created ... not even after thread-sleep 10 seconds after commit or after flushing