[CSDude]

When I tried it a few weeks ago, because ClickHouse names the files based on column names, weird JSON keys resulted in very long filenames and slashes and it did not play well with it the file system and gave errors, I wonder that is fixed?

[setr]

Isn’t that the issue challenge #3 addresses?

https://clickhouse.com/blog/a-new-powerful-json-data-type-fo...

[CSDude]

Tried with the latest version, but it doesn't solve.

    CREATE TABLE mk3
    ENGINE = MergeTree
    ORDER BY (account_id, resource_type)
    SETTINGS allow_nullable_key = 1
    AS SELECT
        *,
        CAST(content, 'JSON') AS content_json
    FROM file('Downloads/data_snapshot.parquet')

    Query id: 8ddf1377-7440-4b4d-bb8d-955cd0f2b723

    ↑ Progress: 239.57 thousand rows, 110.38 MB (172.49 thousand rows/s., 79.48 MB/s.)                                                                                                          22%
    Elapsed: 4.104 sec. Processed 239.57 thousand rows, 110.38 MB (58.37 thousand rows/s., 26.89 MB/s.)

    Received exception:
    Code: 107. DB::ErrnoException: Cannot open file /var/folders/mc/gndsp71j6zz64pm7j2wz_6lh0000gn/T/clickhouse-local-503e1494-c3fb-4a5e-9514-be5ba7940fec/data/default/mk3/tmp_insert_all_1_1_0/content_json.plan.features.available.core/audio.dynamic_structure.bin: , errno: 2, strerror: No such file or directory. (FILE_DOESNT_EXIST)