Execution timeout when loading csv file to ClickHouseContainer

Question

When attempting to upload a CSV file (~100 lines and < 100KB) to ClickHouseContainer running in junit, the client is timing out with the following error:

An exception or error caused a run to abort: Code: 159. Execution timed out, server ClickHouseNode [uri=any://localhost:51733/default]@675749754 

I have the following code to initialize a ClickHouseContainer

val image: ClickHouseContainer =
  new ClickHouseContainer("clickhouse/clickhouse-server:latest")
    .withUrlParam("compress_algorithm", "zstd")
    .withUsername("default")
    .withPassword("")
    .withExposedPorts(8123, 8443, 9000, 9440)
    .waitingFor(
      Wait
        .forHttp("/ping")
        .forPort(8123)
        .forStatusCode(200)
        .withStartupTimeout(Duration.ofSeconds(600)));
image.start()

Once the container starts, I am using ClickHouseClient to load the table from a CSV file.

val server = ClickHouseNode.builder
  .host(image.getHost)
  .port(image.getMappedPort(ClickHouseProtocol.HTTP.getDefaultPort))
  .database("default")
  .credentials(ClickHouseCredentials.fromUserAndPassword("default", ""))
  .build

// Create a CSV file reference
val file = ClickHouseFile.of(csvFile, ClickHouseCompression.NONE, ClickHouseFormat.CSV)

// Create a client
val client = ClickHouseClient.newInstance(server.getProtocol)

// Specify settings, load data to the specified table and wait for completion
client
  .write(server)
  .set("input_format_csv_skip_first_lines", "0")
  .set("format_csv_delimiter", ",")
  .table(tableName)
  .data(file)
  .executeAndWait()

I even tried increasing the socket timeout on the client, but this did not help and instead I was getting connect reset by peer.

val client = ClickHouseClient
  .builder()
  .option(ClickHouseClientOption.SOCKET_TIMEOUT, 10000)
  .option(ClickHouseClientOption.MAX_THREADS_PER_CLIENT, 100)
  .option(ClickHouseClientOption.MAX_EXECUTION_TIME, 1000)
  .build()

The CSV file is too small, hence I don't expect that inserting the rows to clickhouse should take so long.

Answer 1

I can't tell the cause by the information above. But a few things might help with troubleshooting

• Inspect system.query_log and find failed queries associated exception, https://clickhouse.com/docs/en/operations/system-tables/query_log. You can filter by timestamp/query string.
• Increase logger level and inspect logs. You should at least see your queries printed out in the log and error message if insertions fails.
• If you don't see your queries anywhere, consider to rule out container and application environment network, run ClickHouse directly and access via localhost:8123.