horaedb
horaedb copied to clipboard
apache
Apache HoraeDB (incubating) is a high-performance, distributed, cloud native time-series database.
### Describe This Problem The introduction of [opendal](https://github.com/apache/incubator-opendal) as our object_store dependency has been considered [before](https://github.com/CeresDB/ceresdb/issues/608). We would very much welcome someone from the community who can continue to push...
### Describe This Problem In current implementation, there is only one database(called `schema` in src), we should support database creation to allow one server holding multiple databases. ### Proposal Support...
### Describe This Problem Although wal payload in columnar format is supported, the feature is not offered to split the payload in columnar format to avoid large wal log. And...
### Describe This Problem In the current implementation of the storage engine, one table must occupy some resources even if the table may not be used anymore. And the number...
### Describe This Problem Due to the ability to open partition table on multiple ceresdb nodes, there may be inconsistencies in the schema information maintained in memory. Currently, the approach...
### Describe This Problem CeresDB currently handles user error messages unreasonably. When a shard is in the `frozen` state, the error message when reading or writing the tables in this...
### Describe this problem In our deployments with OBKV as wal implementation, we found following errors: ```bash 2023-09-11 10:24:03.322 ERRO [proxy/src/grpc/write.rs:33] Failed to handle write, err:Internal error, msg:Failed to execute...
## Rationale The system table `tables` adds a `partition_info` column to distinguish whether the table is a partition table. ## Detailed Changes Add `partition_info` in `system.public.tables`. ## Test Plan Add...
### Describe This Problem There is a business scenario where tables with the same name exist, and it is hoped that namespace isolation can be done through `database`. ### Proposal...
### Describe This Problem In my local env, I find WAL(rocksdb based) is very large compared with SST, even after manually flush. ``` 1.5G data/wal 76M data/manifest 371M data/store 1.9G...