Use 20s instead.

CHANGELOG.md

@@ -2,7 +2,7 @@
 
 ### Breaking changes
 
-- Client will enable [send_progress_in_http_headers](https://clickhouse.com/docs/en/operations/settings/settings#send_progress_in_http_headers) and set `http_headers_progress_interval_ms` to `29000` (29 seconds) by default. These settings in combination allow to avoid LB timeout issues in case of long-running queries without data coming in or out, such as `INSERT FROM SELECT` and similar ones, as the connection could be marked as idle by the LB and closed abruptly. Currently, 29s is chosen as a safe value, since most LBs will have at least 30s of idle timeout. It can be overridden when creating a client instance if your LB timeout value is even lower than that by manually changing the `send_progress_in_http_headers` value.
+- Client will enable [send_progress_in_http_headers](https://clickhouse.com/docs/en/operations/settings/settings#send_progress_in_http_headers) and set `http_headers_progress_interval_ms` to `20000` (20 seconds) by default. These settings in combination allow to avoid LB timeout issues in case of long-running queries without data coming in or out, such as `INSERT FROM SELECT` and similar ones, as the connection could be marked as idle by the LB and closed abruptly. Currently, 20s is chosen as a safe value, since most LBs will have at least 30s of idle timeout, and, for example, AWS LB sends KeepAlive packets every 20s. It can be overridden when creating a client instance if your LB timeout value is even lower than that by manually changing the `send_progress_in_http_headers` value.
 
   NB: these settings will be enabled only if the client instance was created without setting `readonly` flag (see below).
 

packages/client-common/src/client.ts

@@ -51,16 +51,16 @@ export interface ValuesEncoder<Stream> {
 export type CloseStream<Stream> = (stream: Stream) => Promise<void>
 
 /**
- * By default, {@link send_progress_in_http_headers} is enabled, and {@link http_headers_progress_interval_ms} is set to 290s.
+ * By default, {@link send_progress_in_http_headers} is enabled, and {@link http_headers_progress_interval_ms} is set to 20s.
  * These settings in combination allow to avoid LB timeout issues in case of long-running queries without data coming in or out,
  * such as `INSERT FROM SELECT` and similar ones, as the connection could be marked as idle by the LB and closed abruptly.
- * 290s is chosen as a safe value, since AWS LB timeout is 350s by default and some other LBs might have 300s.
+ * 20s is chosen as a safe value, since most LBs will have at least 30s of idle timeout, and AWS LB sends KeepAlive packets every 20s.
  * It can be overridden when creating a client instance if your LB timeout value is even lower than that.
  * See also: https://docs.aws.amazon.com/elasticloadbalancing/latest/network/network-load-balancers.html#connection-idle-timeout
  */
 const DefaultClickHouseSettings: ClickHouseSettings = {
   send_progress_in_http_headers: 1,
-  http_headers_progress_interval_ms: '290000',
+  http_headers_progress_interval_ms: '20000',
 }
 
 export interface ClickHouseClientConfigOptions<Stream> {

packages/client-node/__tests__/integration/node_client.test.ts

@@ -89,7 +89,7 @@ describe('[Node.js] Client', () => {
     const searchParams = new URL(callURL).search.slice(1).split('&')
     expect(searchParams).toContain('enable_http_compression=1')
     expect(searchParams).toContain('send_progress_in_http_headers=1')
-    expect(searchParams).toContain('http_headers_progress_interval_ms=290000')
+    expect(searchParams).toContain('http_headers_progress_interval_ms=20000')
     expect(searchParams).toContain(jasmine.stringContaining('query_id='))
     expect(searchParams.length).toEqual(4)
   }