You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now the storage engine overwrites duplicate entries, that share the same row key by default. We may also need to support append mode, which never overwrites duplicate entries. However, users might expect to store duplicate rows in the table.
It's helpful to support an append-only mode for the table.
Implementation challenges
We remove duplicate rows in several places.
memtables iters
the time series memtable
the merge tree memtable
the merge reader
during query
during compaction
We have to
add an append_only option and disable dedup in memtables
bypass the merge reader in query and compaction
A remaining problem is how to support DELETE as the merge reader requires each source support dedup. For simplicity, we can only support it while the table is not an append only table. We might need to persist this option in the manifest.
What type of enhancement is this?
API improvement, User experience
What does the enhancement do?
Now the storage engine overwrites duplicate entries, that share the same row key by default. We may also need to support append mode, which never overwrites duplicate entries. However, users might expect to store duplicate rows in the table.
It's helpful to support an append-only mode for the table.
Implementation challenges
We remove duplicate rows in several places.
We have to
append_only
option and disable dedup in memtablesA remaining problem is how to support DELETE as the merge reader requires each source support dedup. For simplicity, we can only support it while the table is not an append only table. We might need to persist this option in the manifest.
Implement History
The text was updated successfully, but these errors were encountered: