kamille

Multiple writer for the same sst caused by `close shard`

2

### Describe this problem Shard will be moved from nodes when process panic because if for any reason, all operations related to such a shard should be stopped before moving(especially...

Tracking Issue: improve compaction

1

### Describe This Problem Now the design of compaction in ceresdb is still so rough, we should make more efforts in it. There are several improvements that can be made...

### Describe This Problem In merge iterator, we extract row by row from the src record batch, and append row by row to the dst record batch now. However, arrow...

### Describe This Problem When I run tsbs with the default memtable size(30M), I found the flushed sst is too small(1.9M). As inspecting the parquet file, the high compression ratio...

Shard based recovery

1

### Describe This Problem Now table recovery in on table level but wal's storing in on shard level. The recovery performance may be unsatisfied especially in kafka based wal. ###...

### Describe This Problem # Data organization When using the primary key `tsid + timestmap`, the data will be ordered like: ``` tag1, tag2, tag3, timestamp1, field1 tag1, tag2, tag3,...

### Describe This Problem The `flushed` point in table version is updated before the `deleted` point's updating in wal manager. ``` // Edit table version to remove dumped memtables. let...

### Describe This Problem As I see, write test related to the structs mentioned in title is too annonying. It make me mad to find the way of building them....

General wal deletion model

3

### Describe This Problem In fact, I want to refactor the whole wal module finally, reasons: + hard to implement master-follower model based on current wal implementation. + hard to...

**Describe this problem** See title. **Steps to reproduce** + let table dropped in ceresdb but not dropped in meta. + write to the table in the same shard as above...