Samza has pretty good support for collectng metrics: https://samza.apache.org/learn/documentation/0.10/container/metrics.html
With these metrics logged in Kafka, you can simply consume from this stream for monitoring. In our case, we pipe these metrics into OpenTSDB for visualization. For example, you can monitor the rate of processing by a specific container (msgs/sec) by looking at SamzaContainerMetrics.process-envelopes. Consumer lag is also an important metric, which is named kafka-*-*-messages-behind-high-watermark. You will need to take a peek at the published JSON to get a sense of what metrics are reported. Thanks, David On Mon, Aug 8, 2016 at 12:50 AM 李斯宁 <lisin...@gmail.com> wrote: > Hi guys, > Is there a simpler way for monitoring: "Is every task in my samza jobs are > normally running and making progress?" > > Currently, the samza job's Application Master's web do not show the > information in a easy-to-discover way. If you want information of each > task, you need to click to view every container's log or connect to every > container's jmxport to get information. > > I already found that we can do some monitoring from 'coordinate' and > 'checkpoint' streams, but > is there a tool available for such monitoring requirement? I think it is a > very common need. > -- > 李斯宁 >
