避免发生大查询热身的活动 [英] Activities to avoid big query warm-up occurence

问题描述

在流数据文章中，提到几个小时不活动后，下一次插入将再次发生预热期." https://developers.google.com/bigquery/streaming-data-into-bigquery

On the streaming data article, it is mentioned "After several hours of inactivity, the warm-up period will occur again during the next insert." https://developers.google.com/bigquery/streaming-data-into-bigquery

哪些活动可以使连接保持温暖? 我将编写连接池，预计它将提供bigquery对象，该对象可以在没有任何预热期的情况下放入数据.

What are those activities that can keep the connection warm? I'll be writing connection pool and it is expected to provide bigquery object that can put data in without any warm up period.

推荐答案

流文档中讨论的预热时间与查询中是否可以看到最近流到表的记录有关，而不是与表是否将立即接受流记录有关. .如果某个表是新表，或者已经有几个小时没有通过流接收记录，那么在生成查询结果时，可能需要2分钟的时间才能对这些新流记录进行扫描.

The warmup period discussed in the streaming documentation is related to whether recently streamed records to a table are made visible to queries, not whether the table will accept streaming records immediately. If a table is new or hasn't been receiving records via streaming for hours, then it may take up to 2 minutes before these freshly streamed records are scanned when generating query results.

除非您的使用模式是每隔几个小时流一次记录然后立即查询，否则不必担心.在这种情况下，即使每小时只有少量流记录也足以使表热身".

Unless your usage pattern is to only stream records every few hours and then immediately query, this should not be a concern. Even a small number of streaming records per hour is enough to keep the table "warmed up" in this context.

这篇关于避免发生大查询热身的活动的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！