datastaxagent_backup_local_disk

#!/usr/bin/env bash

#
# datastaxagent_backup_local_disk
#
# Script to recursively backup the cassandra data directory.
# Uses rsync to remove old files and copy only new files.
# License - Apache 2.0 
#

#
# TODO - rsync -e ssh for security. Courtesy Matt Atwater
# TODO - don't backup the live table folder then remove the "errcode=0; ;;" 2 uses below.
# TODO - Capture list of keyspaces
# TODO - Capture each keyspace > Keyspace_name_schema.cql
# TODO - Capture hostname, hostname -i, date
#
#
# Planned change
# Compactions and write path continuously change the files to be rsynced. Rsync detects this and after completing the backup, returns errors which I replace with a 0. See
# 
# 	23) echo     Partial transfer due to error; errcode=0; ;;
# 	24) echo     Partial transfer due to vanished source files; errcode=0; ;;
#
# The script was always intended to be extended to backup only the snapshot directories.
# The change would look like:
#
#       rsync -avzhH --delete */snapshot "$backup_path"
# and 
#
#	23) echo     Partial transfer due to error; ;;
#	24) echo     Partial transfer due to vanished source files;  ;;
#

# allow invokation of related scripts in this dir
scriptdir=`dirname "$BASH_SOURCE"`
PATH="$scriptdir:$PATH"

# Directory to backup
source_backup_path=/mnt/cassandra

# List of backup destination paths
declare -A backup_path_list
backup_path_list[0]=/backup1/cassandra
backup_path_list[1]=/backup2/cassandra
backup_path_count=${#backup_path_list[@]}

# Choose backup destination path by modulo of date since Jan 1 1970
backup_start=$(date)
date=$(date +%F)
epoch_seconds=$(date '+%s')
epoch_day=$(echo $epoch_seconds '/60/60/24' | bc -q )
backup_path_number=$(echo $epoch_day "%$backup_path_count" | bc -q )
backup_path=${backup_path_list[$backup_path_number]}
backup_log="$backup_path/log.$date"
backup_path_files="$backup_path/files"

# Check and create directories
if [ x"$backup_path" == x ]; then echo Error: Missing backup path "$backup_path" ; exit 254; fi
if [ ! -d "$backup_path" ]; then echo Error: Backup path, "$backup_path", is not a directory; exit 253; fi
if [ ! -d "$backup_path_files" ]; then
	mkdir  "$backup_path_files"
	chown cassandra:cassandra "$backup_path_files"
	chmod 777 "$backup_path_files"
fi

exec >> "$backup_log" 2>&1   # send all output to the log
echo ''
echo backup date $backup_start
echo backup_path_count $backup_path_count
echo seconds since epoch: $epoch_seconds
echo days since epoch: $epoch_day
echo backup path number: $backup_path_number
echo backup path: "$backup_path"
if [ ! -d "$source_backup_path" ]; then echo Error: backup path "$source_backup_path" is not accessible; exit -252; fi
echo ''

# Copy files to backup
cat > /dev/null 2>&1   #read and discard the directory names

echo backing "$source_backup_path"  to "$backup_path_files"
errcode=99
rsync -avzhH --delete "$source_backup_path" "$backup_path_files"
errcode=$?
printf '%s %s - ' errcode $errcode
case $errcode in
	 0) echo     Success; ;;
	 1) echo     Syntax or usage error; ;;
	 2) echo     Protocol incompatibility; ;;
	 3) echo     Errors selecting input/output files\, dirs ;;
	 4) echo     Requested action not supported: an attempt was made to manipulate 64-bit files on a platform that cannot support them or an option was specified that is supported by the client and not by the server.; ;;
	 5) echo     Error starting client-server protocol; ;;
	 6) echo     Daemon unable to append to log-file; ;;
	10) echo     Error in socket I/O; ;;
	11) echo     Error in file I/O; ;;
	12) echo     Error in rsync protocol data stream; ;;
	13) echo     Errors with program diagnostics; ;;
	14) echo     Error in IPC code; ;;
	20) echo     Received SIGUSR1 or SIGINT; ;;
	21) echo     Some error returned by waitpid\(\); ;;
	22) echo     Error allocating core memory buffers; ;;
	23) echo     Partial transfer due to error; errcode=0; ;;
	24) echo     Partial transfer due to vanished source files; errcode=0; ;;
	25) echo     The --max-delete limit stopped deletions; ;;
	30) echo     Timeout in data send/receive; ;;
	35) echo     Timeout waiting for daemon connection; ;;
	*) echo unknown status; ;;
esac
if [ $errcode == 0 ]; then
	echo Backup successful. Started $backup_start. Completed $(date). ; exit $errcode
else
	echo Backup failed: Started $backup_start. Aborted $(date). ; exit $errcode;
fi