Description of problem: Agent loads the Sigar library with nonative flag activated. This leads to errors agent startup failure in cases where the Sigar library is not compatible with the operating system. How reproducible: Every time Steps to Reproduce: 1. Use a system that is not compatible with Sigar (eg. SUSE 10.1) 2. Run the agent with --nonative flag 3. Notice error Actual results: Failure to start with the following error: # # A fatal error has been detected by the Java Runtime Environment: # # SIGFPE (0x8) at pc=0x00002b592b3d668f, pid=13696, tid=1074792768 # # JRE version: 7.0_07-b10 # Java VM: Java HotSpot(TM) 64-Bit Server VM (23.3-b01 mixed mode linux-amd64 compressed oops) # Problematic frame: # C [ld-linux-x86-64.so.2+0x868f] do_lookup_x+0xcf # # Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again Expected results: The agent starts without Sigar loaded Additional info: This issue has been observed on Suse 10.x 64bit. This version of Suse is running an older glibc library (2.4.x) that should be compatible with Sigar but it is throwing SIGPFE excetion at startup. The main problem is that --nonative flag is not working properly as it should have prevented the loading of Sigar jars completely.
Moved static block of code that was checking for Sigar availability to a static method. This allows correct usage of system properties and stored agent preferences at startup. With this change the agent will not load the Sigar library even if --nonative flag or rhq.agent.disable-native-system property are set to true.
master branch commit: http://git.fedorahosted.org/cgit/rhq/rhq.git/commit/?id=cb654429a9c0d0d30092a21740da5e1d4e614be2
Bulk closing of issues in old RHQ releases that are in production for a while now. Please open a new issue when running into an issue.