[IMPALA-160] Impala crashes on startup on AMD machines without ssse3 support.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: Impala 0.6
• Fix Version/s: Impala 1.2.1
• Labels: SIGILL crash impala
• Environment: IMPALA 0.6-1.p0.109, CDH 4.2.0-1.cdh4.2.0.p0.10 on Ubuntu 12.04.2 LTS, Hadoop cluster created with Cloudera Manager free 4.5.0 using parcels
• Assignee: Nong Li
• Reporter: Vladimir Nicolici
• Created: 20/Mar/13 19:44
• Updated: 16/Mar/17 03:42
• Resolved: 09/Jan/14 21:32

Description

When attempting to create a Cloudera Hadoop cluster including Impala, or when attempting to add an Impala service to an existing cluster, the operation fails with this error:

Service did not start successfully; not all of the required roles started: No Impalads are running.

The beginning of stdout for each impalad instance shows the error that causes the crash:

# A fatal error has been detected by the Java Runtime Environment:
#
# SIGILL (0x4) at pc=0x00000000006a22ca, pid=11285, tid=140297147213632
#
# JRE version: 6.0_31-b04
# Java VM: Java HotSpot(TM) 64-Bit Server VM (20.6-b01 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C [impalad+0x2a22ca] boost::uuids::detail::sha1::process_block()+0xda
#
# An error report file with more information is saved as:
# /run/cloudera-scm-agent/process/342-impala-IMPALAD/hs_err_pid11285.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.

The strange thing is I previously installed two other test clusters using the same software stack and architecture, and I didn't hit this problem.

But in this case, I even formatted the hosts and reinstalled everything including the OS, and I hit the same problem again.

I attached the directory with the logs and the inspector output for the cluster.