Separate JVM running Daily Stats Aggregation failed to start due to insufficient Stack Size (Bad rows processed).

book

Article ID: 48581

calendar_today

Updated On:

Products

CA Application Performance Management Agent (APM / Wily / Introscope) INTROSCOPE

Issue/Introduction

Description:

When the daily stats aggregation does not occur due to its JVM failing to start, you may see an error message in the EM log such as:

[WARN] [DailyAggregation.Thread1] [Manager.com.timestock.tess.services.processors.StatsAggregator] Bad rows processed:

The typical/default value of "-XX:ThreadStackSize" is 128 and is sufficient for most JVMs. Occasionally, this stack size is too small for some Windows and Linux JVMs.

Solution:

To work around this issue, depending on the JVM bit mode, perform one of the following tasks on the EM running the Stats Aggregation

32-bit JVMs:

  • Stop EM

  • Edit the file <em-root-dir>/config/tess-default.properties and modify dailystats.jvmArgs to increase the "-XX:ThreadStackSize" to 192. If this stack size is still not sufficient, increase it to 256. The default value of "-XX:ThreadStackSize" has been increased to 192 in release 9.1.7 and future release 9.5.1.

  • Start EM

64-bit JVMs:

  • Stop EM

  • Limiting the thread stack is not necessary because the heap on a 64-bit JVM can be many times what it can be on a 32-bit JVM. Therefore edit the file <em-root-dir>/config/tess-default.properties and modify dailystats.jvmArgs to remove the "-XX:ThreadStackSize" parameter.

  • Start EM

Environment

Release:
Component: APMCM