From 393c2ec93b261db90aa3c258743f8225863fdab3 Mon Sep 17 00:00:00 2001
From: tim <tim@treehouse.org.za>
Date: Thu, 31 Aug 2017 16:41:28 +0200
Subject: [PATCH] Adds a script to verify data checksums and notify of change
 and potential corruption, using cshatag (or shatag).

---
 verify_data.sh | 72 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 72 insertions(+)
 create mode 100755 verify_data.sh

diff --git a/verify_data.sh b/verify_data.sh
new file mode 100755
index 0000000..b4a3710
--- /dev/null
+++ b/verify_data.sh
@@ -0,0 +1,72 @@
+#!/bin/bash
+
+# Compute checksums for important data, and report changed or invalid 
+# checksums.
+# This should help to identify bit rot among rarely-used files.
+
+mailto=root@localhost
+
+if ! [ $( which cshatag ) ]; then
+	echo "cshatag is not installed. Exiting."
+	exit
+fi
+
+if ! [ $( which pcre2grep ) ]; then
+	echo "pcre2grep is not installed. Exiting."
+	exit
+fi
+
+shatag="cshatag"
+
+for dir in /data/www /data/home ; do
+
+	output=$( find "${dir}" -xdev -type f -exec ${shatag} {} \; )
+	final_output=
+	corrupt_output=$( echo "${output}" | grep -v "<ok>" | pcre2grep -vM '<outdated>.*\n.*stored:.*\n.*actual.*' )
+	outdated_output=$( echo "${output}" | grep -v "<ok>" | pcre2grep -vM '<corrupt>.*\n.*stored:.*\n.*actual.*' )
+	subject="Changed files in ${dir}"
+
+	if [ -n "${corrupt_output}" ]; then
+		final_output=$( cat <<-END_CORRUPT
+			${final_output}
+
+			The following files may be corrupted:
+
+			${corrupt_output}
+
+		END_CORRUPT
+		)
+		subject="CORRUPT files in ${dir}"
+	fi
+	if [ -n "${outdated_output}" ]; then
+		modified_output=$( echo "${outdated_output}" | pcre2grep -vM '<outdated>.*\n.*stored:.*0000000000.000000000.*\n.*actual.*' )
+		final_output=$( cat <<-END_OUTDATED
+			${final_output}
+
+			The following files have been modified:
+
+			${modified_output}
+
+		END_OUTDATED
+		)
+	fi
+
+	if [ -n "${final_output}" ]; then
+	       cat <<-END_FINAL | mail -s "${subject}" "${mailto}"
+The following files have changed on disk.
+
+If this is unexpected (that is, you haven't opened and re-saved these
+files), they may have become corrupted. Please forward this list to your
+systems administrator and request they be restored from backups.
+${final_output}
+		END_FINAL
+	fi
+done
+
+# for shatag:
+#	db=$( echo $dir | sed 's/[\/ ]/-/g' | sed 's/^-//' )
+#	output=$( find "${dir}" -type f -exec ${shatag} -d "/root/${db}.shatagdb" -vqpt {} + 2>&1 )
+#	outdated_output=$( echo "${output}" | grep -v "<missing>" )
+#
+#	corrupt_output=$( echo "${output}" | grep -v "<ok>" | awk -v SKIP=-1 '/<outdated>/ { SKIP = 2 } SKIP-- >= 0 {next } 1' )
+#	outdated_output=$( echo "${output}" | grep -v "<ok>" | awk -v SKIP=-1 '/<corrupt>/ { SKIP = 2 } SKIP-- >= 0 {next } 1' )