大网数据广告活动
服务器运维问答

Monitoring and alerting on Google Cloud Bigtable

· 52CE 我爱测

We are building a data pipeline on GCP and chosen Bigtable as our data warehouse for raw events. My problem is - I cannot find a way to properly monitor the cluster. The dashboard offers a limited amount of metrics which could do…

问题原文

原帖链接 · Yaron Idan · CC BY-SA 3.0

We are building a data pipeline on GCP and chosen Bigtable as our data warehouse for raw events. My problem is - I cannot find a way to properly monitor the cluster. The dashboard offers a limited amount of metrics which could do the trick if there was any way to connect them into StackDriver without too much hassle, just like all the other services on GCP are set up.

Is it possible to monitor Bigtable with Stackdriver? If not, is there another recommended way to achieve this?

I found this post that explains how to get reports on the requests running against Bigtable, but this doesn't help me when trying to receive the big picture about the entire cluster's health.

已采纳解答

原帖链接 · Misha Brukman · CC BY-SA 4.0

Google Cloud Bigtable exports server-side metrics to Stackdriver; you can find more information in the Bigtable documentation as well as Stackdriver metrics list.

In addition, if you are using the HBase-compatible Java client for Bigtable, you can export Dropwizard metrics on the client side which will give you detailed client-side performance metrics, such as end-to-end request latency, that you can track on your own, e.g., via your own Graphite instance.

These metrics are not meant to be a substitute to Stackdriver metrics for Bigtable, but to complement those metrics, and we would recommend tracking client-side metrics in addition to looking at server-side metrics.

摘自公开归档,保留原文语言,版式经整理,外部媒体与邮箱地址已移除。

资料来源与适用范围

资料来源
Server Fault
适用范围
以原帖系统、软件版本与发布日期为准(2017-01-01)
内容语言
中文;命令、路径和错误标识保留技术原文
本地收录日期
2026-11-15
授权说明
CC BY-SA 3.0 / CC BY-SA 4.0

正文与所需资源由本站本地存储。查看来源及更新记录 ↗

请按适用版本和实际日志确认原因;本站没有对每一种系统或设备逐项执行文中的操作。