Files
impala/testdata/bin/run-hbase.sh
Michael Brown 6d8c075d9c IMPALA-1996: Start HBase per directions in documentation; Implement HBase startup retry
I. Start HBase per directions

1. https://hbase.apache.org/book.html#_configuration_files mentions a
'regionservers' file that points to a list of hosts on which to start
HBase RegionServers. When HBase starts in our mini-cluster there are
messages printed like this:

cat: /home/mikeb/Impala/fe/src/test/resources/regionservers: No such file or directory

The presence of this file now starts a single RegionServer and takes the
place of RegionServer 1 in the "additional region servers" startup, a
separate call.

2. The additional RegionServers are started but now we only start 2 from
index 2. See https://hbase.apache.org/book.html#quickstart_pseudo

There are still 3 total RegionServers using the same ports as before. We
are simply configuring our settings as directed in the documentation.

There were mentions in testdata/bin/run-hbase.sh of a "hbase race". One
possible such bug is https://issues.apache.org/jira/browse/HBASE-5780
which has been fixed for a while. I've removed the check to wait for
that Master, though I have not removed the Python script that does the
waiting. We could remove that later after we let this patch bake.

Also, https://issues.apache.org/jira/browse/HBASE-4467 has been marked
"not a problem", so I've removed references to that.

II. Implement HBase start retry

If starting either HBase Master or additional RegionServers fails, kill
all of HBase and try again.  Do this for some number of attempts.

In order to keep errexit ("set -e") happy, we expect the possibility of
some of the startup attempts failing. We use control flow in those
cases. In the last case, errexit can fail on our behalf.

There is some code duplication here, but because Bash can't give us a
stack trace on failure, and only a line number, I chose not to use
functions to handle reuse. We don't really have functions anywhere else
at the moment, either.

Testing:

It's pretty difficult to try to trigger a real "HBase fails to start"
situation. I tested my changes by faking HBase failures, both when
starting up the Master and first RegionServer, and also starting
subsequent RegionServers.

Multiple private builds have passed.

Change-Id: Ib1d055a8a9098ce24e2f31b969501b6e090eab19
Reviewed-on: http://gerrit.cloudera.org:8080/2804
Reviewed-by: Michael Brown <mikeb@cloudera.com>
Tested-by: Internal Jenkins
2016-05-12 14:17:33 -07:00

114 lines
3.8 KiB
Bash
Executable File

#!/bin/bash
# Copyright (c) 2012 Cloudera, Inc. All rights reserved.
set -euo pipefail
trap 'echo Error in $0 at line $LINENO: $(cd "'$PWD'" && awk "NR == $LINENO" $0)' ERR
CLUSTER_BIN=${IMPALA_HOME}/testdata/bin
HBASE_JAAS_CLIENT=${HBASE_CONF_DIR}/hbase-jaas-client.conf
HBASE_JAAS_SERVER=${HBASE_CONF_DIR}/hbase-jaas-server.conf
HBASE_LOGDIR=${IMPALA_CLUSTER_LOGS_DIR}/hbase
# Kill and clean data for a clean start.
${CLUSTER_BIN}/kill-hbase.sh > /dev/null 2>&1
# Gives HBase startup the proper environment
cat > ${HBASE_CONF_DIR}/hbase-env.sh <<EOF
#
# This file is auto-generated by run-hbase.sh. Do not edit.
#
export JAVA_HOME=${JAVA_HOME}
export HBASE_LOG_DIR=${HBASE_LOGDIR}
export HBASE_PID_DIR=${HBASE_LOGDIR}
EOF
# Put zookeeper things in the logs/cluster/zoo directory.
# (See hbase.zookeeper.property.dataDir in hbase-site.xml)
rm -rf ${IMPALA_CLUSTER_LOGS_DIR}/zoo
mkdir -p ${IMPALA_CLUSTER_LOGS_DIR}/zoo
mkdir -p ${HBASE_LOGDIR}
if ${CLUSTER_DIR}/admin is_kerberized; then
#
# Making a kerberized cluster... set some more environment
# variables and other magic.
#
. ${MINIKDC_ENV}
if [ ! -f "${HBASE_JAAS_CLIENT}" ]; then
echo "Can't find ${HBASE_JAAS_CLIENT}"
exit 1
fi
if [ ! -f "${HBASE_JAAS_SERVER}" ]; then
echo "Can't find ${HBASE_JAAS_SERVER}"
exit 1
fi
# Catch the case where the /hbase directory is not owned by the
# hbase user. This can happen when the cluster was formed without
# kerberos and then remade with "create-test-configuration.sh -k".
if HBASE_LS_OUTPUT=`hadoop fs -ls -d /hbase 2>&1`; then
if echo ${HBASE_LS_OUTPUT} | tail -n 1 | grep -q -v " hbase "; then
# /hbase not owned by 'hbase'. Failure.
echo "The HDFS /hbase directory is not owned by \"hbase\"."
echo "This can happen if the cluster was created with kerberos,"
echo "and then switched to kerberos without a reformat."
fi
fi
# These ultimately become args to java when it starts up hbase
K1="-Djava.security.krb5.conf=${KRB5_CONFIG}"
K2="${JAVA_KRB5_DEBUG}"
K3="-Djava.security.auth.login.config=${HBASE_JAAS_CLIENT}"
K4="-Djava.security.auth.login.config=${HBASE_JAAS_SERVER}"
# Add some kerberos things...
cat >> ${HBASE_CONF_DIR}/hbase-env.sh <<EOF
export HBASE_OPTS="${K1} ${K2} ${K3}"
export HBASE_MANAGES_ZK=true
export HBASE_ZOOKEEPER_OPTS="${K1} ${K2} ${K4}"
export HBASE_MASTER_OPTS="${K1} ${K2} ${K4}"
export HBASE_REGIONSERVER_OPTS="${K1} ${K2} ${K4}"
EOF
fi
: ${HBASE_START_RETRY_ATTEMPTS=5}
# `rm -f` hbase startup output capture so that `tee -a` below appends
# only for the lifetime of this script
rm -f ${HBASE_LOGDIR}/hbase-startup.out ${HBASE_LOGDIR}/hbase-rs-startup.out
for ((i=1; i <= HBASE_START_RETRY_ATTEMPTS; ++i)); do
echo "HBase start attempt: ${i}/${HBASE_START_RETRY_ATTEMPTS}"
echo "Killing any HBase processes possibly lingering from previous start attempts"
${IMPALA_HOME}/testdata/bin/kill-hbase.sh
if ((i > 1)); then
HBASE_WAIT_AFTER_KILL=$((${i} * 2))
echo "Waiting ${HBASE_WAIT_AFTER_KILL} seconds before trying again..."
sleep ${HBASE_WAIT_AFTER_KILL}
fi
if ((i < HBASE_START_RETRY_ATTEMPTS)); then
# Here, we don't want errexit to take effect, so we use if blocks to control the flow.
if ! ${HBASE_HOME}/bin/start-hbase.sh 2>&1 | tee -a ${HBASE_LOGDIR}/hbase-startup.out
then
echo "HBase Master startup failed"
elif ! ${HBASE_HOME}/bin/local-regionservers.sh start 2 3 2>&1 | \
tee -a ${HBASE_LOGDIR}/hbase-rs-startup.out
then
echo "HBase regionserver startup failed"
else
break
fi
else
# In the last iteration, it's fine for errexit to do its thing.
${HBASE_HOME}/bin/start-hbase.sh 2>&1 | tee -a ${HBASE_LOGDIR}/hbase-startup.out
${HBASE_HOME}/bin/local-regionservers.sh start 2 3 2>&1 | \
tee -a ${HBASE_LOGDIR}/hbase-rs-startup.out
fi
done
echo "HBase startup scripts succeeded"