李洁 2013-02-17, 03:28
邵赛赛 2013-02-18, 01:09
李洁 2013-02-19, 15:53
Since aggregation script will aggregate per-node data and put the data into HBase table “ClusterSummary”, so could you please check this table to see whether there is any data in this table.
Also HICC fetch data from HBase table using REST through HICC widget configuration descriptor, you will know how HICC get specific data by which table, which column family and qualify through checking these descriptors. Also you can add some additional widgets like: System Metrics, host selector, cluster selector and others to your web page to test if HICC can fetch data from HBase, this can minimize your troubleshooting scale.
You can see ChukwaHBaseStore.java to see the details of retrieving data by HICC.
Lastly make sure to check logs to see whether there is any related exception.
From: 李洁 [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, February 19, 2013 11:53 PM
To: [EMAIL PROTECTED]
Subject: Re: chukwa0.5.0配置问题
Thank you for your answer, I forgot to say I had run cluster aggregation script,and collector thread can write data to Hbase ,
But I don't know whether the data in Hbase table is correct, How do I check the correctness of data,Do you have any other information or ideas? Looking forward to your answer.
2013/2/18 邵赛赛 <[EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>>
The NullPointException you metioned in your doc file has already been recovered in Chukwa 0.6.0, you can check out the trunk code and try it.
For the problem of no data display in HICC web page, could you please check the follow steps:
1. Check the HBase table SystemMetrics whether HBase store correct data.
2. As the web page display cluster aggregated data, you should install and use Pig to aggregate your collected per-node data periodically, for the details you can see Chukwa 0.5 quick start guide.
Ps. would you please use English, so other guys who cannot understand Chinese can help you.[cid:[EMAIL PROTECTED]7150]
在 2013年2月17日上午11:28，李洁 <[EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>>写道：