Failed to save modifications to the server. Error returned: 'OLE DB or ODBC error: [DataSource.Error] ERROR [HY000] Incomplete input stream, expected at least 28 more bytes.

[Alexander Fokin]

Hi! I've got an error when I try to get data to Power BI from Clickhouse through ODBC connector. It can be seen that the data starts to load, but after a few seconds this error appears:
Failed to save modifications to the server. Error returned: 'OLE DB or ODBC error: [DataSource.Error] ERROR [HY000] Incomplete input stream, expected at least 28 more bytes.

Timeout ODBC = 900000

Query is really simple, like:
SELECT *
FROM scheme1.table1
LIMIT 1000000

scheme1.table1 has about 15 mln of rows.

Could you suggest what is the reason of that issue?

[VLADIMIR KOBZEV]

Have the same problem!

[d4rkr4in]

So annoying, the same issue. It can sometimes happen sometimes not.

[d4rkr4in]

Hey guys, try to NOT use SQL Query in Power Query. Use native buttons for that. Get Source - Additional - ODBC - OK (don't write any SQL query).
It should help you. Try and get feedback

[GrabsOrion]

Same problem, but with Informatica PowerCenter.

[nikerov-kirill]

Same problem. Error OLE DB or ODBC: [DataSource.Error] ERROR [HY000] Incomplete input stream, expected at least 201 more bytes. Does anyone have a solution?

[akzhar]

Same problem.
Error: PipelineException: ERROR [HY000] Incomplete input stream expected at least 4 more bytes
Have no idea what's the reason... It sometimes happen sometimes not.

[vladdell]

Same problem with SSAS Tabular.

[pauliusmiskunas]

Similar problem here, trying to get data to Qlik Sense from Clickhouse through ODBC connector.

The following error occurred:
QVX_UNEXPECTED_END_OF_DATA: SQL##f - SqlState: HY000, ErrorCode: 1, ErrorMsg: Incomplete 
input stream, expected at least 1192 more bytes

It can sometimes happen sometimes not. When load is finished without error, the number of rows doesn't match the table rows number in database (not all rows are loaded), also getting different number of records each load.
ODBC version - Release 1.2.1.20220905
ClickHouse - 24.3.2.23
table DDL:

CREATE TABLE db_name.table_name
(
    `id` UInt64,
    `title` Nullable(String),
    `item_price` Nullable(Float64),
    `description` Nullable(String),
    `image_url` Nullable(String),
    `product_url` Nullable(String),
    `manufacturer_code` Nullable(String),
    `article_number` Nullable(String),
    `sale_price` Nullable(Float64),
    `manufacturer` Nullable(String),
    `short_message` Nullable(String),
    `specs` Nullable(String),
    `brand` Nullable(String),
    `categories` Nullable(String),    
    `margin_priority` Nullable(String),
    `eans_values` Nullable(String),
    `updated_at` DateTime
)
ENGINE = MergeTree
PARTITION BY toYYYYMM(updated_at)
ORDER BY id
SETTINGS index_granularity = 8192;

[Paulo Rijnberg]

I am getting the same error. For some context:

• I got a fact model with ~ 26M records.
• The error only occurs when I add an additional column via Power Query. The column is simply multiplying the units with the price.
• I am currenly using the import mode, since we are stumbling upon a Clickhouse 500 ERROR when creating relationships between fact tables and bridge tables (we are following Kimballs architecture).

When adding the given column I get the following error:

Found a workaround; I noticed when adding the column via DAX, the error doesn't appear. Note, I think adding a column via DAX might sometimes be less efficient. Also, I don't think this column is included in the data model but only created in the report. So if you are creating a single data model with multiple dashboards, you might need to add the DAX column multiple times.

Hopefully this will be fixed soon!

[Eugene Klimov]

try to execute
in clickhouse-client
SELECT ... FORMAT Null
maybe issue is not related to ODBC and related to your selected tables data