The variance for this alert is 1,091,249,401B which is 37% of the 2,971,031,562B average (594,206,312B is the limit)

[#1]

Getting an Alert of Variance on HDFS service in HDP

The variance for this alert is 1,091,249,401B which is 37% of the 2,971,031,562B average (594,206,312B is the limit)

[#2]

@nshelke

Can you please share the complete Alert stack (if it is WARN/CRITICAL)?

Usually the variance for this alert shows the alert about how quick your namenode heap usage is growing.

You can check the following link to findout if you have tuned the NameNode Heap size properly based on the number of files?

https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.4.2/bk_installing_manually_book/content/ref-809...

.

[#3]

@Jay SenSharma

Please find the attached logs.PFAambari-alerts.txt

[#4 · Accepted solution]

@nshelke

These warning messages indicates that the growth rate was higher than the one that is defined in the alert growth rate threshold in a specified day/week. Normally this happens when we run some excessive job or perform load test to store large data on HDFS. We can tune the threshold based on our requirement or based on our growth observations by clicking on the "Edit" button on the alert definition on the Ambari UI.

[WARNING] [HARD] [HDFS] [increase_nn_heap_usage_daily] (NameNode Heap Usage (Daily)) The variance for this alert is 63MB which is 34% of the 186MB average (37MB is the limit)

This service-level alert is triggered if the NameNode heap usage deviation has grown beyond the specified threshold within a day period.

.

[WARNING] [HARD] [HDFS] [namenode_increase_in_storage_capacity_usage_daily] (HDFS Storage Capacity Usage (Daily)) The variance for this alert is 950,843,960B which is 36% of the 2,626,832,493B average (788,049,748B is the limit)

This service-level alert is triggered if the increase in storage capacity usage deviation has grown beyond the specified threshold within a day period."

.