PT-1336 pt-stalk removes user files from the destination directory

* Added check that files to be removed were created by pt-stalk

* Added option -maxdepth 1 to find command, so the tool does not search for files in the nested directories

* Added unit tests

* This should fix PT-1812 also

bin/pt-stalk

@@ -1351,13 +1351,17 @@ sleep_ok() {
 }
 
 purge_samples() {
-   local dir="$1"
+   local dir=`realpath -s "$1"`
    local retention_time="$2"
    local retention_count="$3"
    local retention_size="$4"
 
    # Delete collect files which more than --retention-time days old.
-   find "$dir" -type f -mtime +$retention_time -exec rm -f '{}' \;
+   if [ -n "$OPT_PREFIX" ]; then
+      find "$dir" -maxdepth 1 -type f -mtime +$retention_time -name "$OPT_PREFIX-*" -exec rm -f '{}' \;
+   else
+      find "$dir" -maxdepth 1 -type f -mtime +$retention_time -regextype posix-egrep -regex "$dir/[0-9]{4}(_[0-9]{2}){5}-.*" -exec rm -f '{}' \;
+   fi
 
    local oprofile_dir="/var/lib/oprofile/samples"
    if [ -d "$oprofile_dir" ]; then
@@ -1370,21 +1374,21 @@ purge_samples() {
    targetCnt=$(($retention_count + 0))
    if [ $targetCnt -gt 0 ]; then
       targetCnt=$(($retention_count + 1))
-      files_to_delete=$(find $dir -type f -exec basename {} \; | cut -f1 -d- | sort -r | uniq | tail -n +${targetCnt})
+      files_to_delete=$(find "$dir" -maxdepth 1 -type f -exec basename {} \; | cut -f1 -d- | sort -r | uniq | tail -n +${targetCnt})
       for prefix in $files_to_delete; do
-          echo "deleting files ${dir}${prefix}* according to the --retention-count param"
-          rm -f ${dir}${prefix}* 2>/dev/null
+          echo "deleting files ${dir}/${prefix}* according to the --retention-count param"
+          rm -f ${dir}/${prefix}* 2>/dev/null
       done
    fi
 
    targetSize=$(($retention_size + 0))
    if [ $targetSize -gt 0 ]; then
-      files_to_delete=$(find $dir -type f -exec basename {} \; | cut -f1 -d- | sort -r | uniq | tail -n +1)
+      files_to_delete=$(find "$dir" -maxdepth 1 -type f -exec basename {} \; | cut -f1 -d- | sort -r | uniq | tail -n +1)
       for prefix in $files_to_delete; do
           current_size=$(du -BM $dir | cut -f1 -d"M")
           if [ $current_size -gt $targetSize ]; then
-             echo "deleting files ${dir}${prefix}* according to the --retention-size param"
-             rm -f ${dir}${prefix}* 2>/dev/null
+             echo "deleting files ${dir}/${prefix}* according to the --retention-size param"
+             rm -f ${dir}/${prefix}* 2>/dev/null
           else
              break
           fi

sandbox/test-env

@@ -295,6 +295,7 @@ fi
 case $opt in
     start)
         cd $PERCONA_TOOLKIT_BRANCH/sandbox
+        echo "${2:-"master"}" 12345
         ./start-sandbox "${2:-"master"}" 12345
         exit_status=$((exit_status | $?))
         set_mysql_version

t/pt-stalk/pt-stalk.t

@@ -422,7 +422,140 @@ like(
    qr/matched=yes/,
    "Accepts floating point values as treshold variable"
 );
+
+# ###########################################################################
+# Variable declaration for the retention tests
+# ###########################################################################
+
+my $odate;
+
+# ###########################################################################
+# Test if retention does not remove files that were not collected
+# ###########################################################################
+
+cleanup();
+
+system("mkdir $dest");
+$odate=`date --rfc-3339=date --date='-3 month'`;
+system("touch -d '$odate' $dest/nostalk");
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --prefix nostalk --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = `ls -l $dest`;
+
+like(
+   $output,
+   qr/nostalk/m,
+   "Retention test 1: Not-matched file not touched"
+);
           
+# ###########################################################################
+# Test if files that match the prefix-, are removed by the retention option
+# ###########################################################################
+
+cleanup();
+
+system("mkdir $dest");
+$odate=`date --rfc-3339=date --date='-3 month'`;
+system("touch -d '$odate' $dest/nostalk-");
+system("touch -d '$odate' $dest/nostalk-innodbstatus1");
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --prefix nostalk --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = `ls -l $dest`;
+
+unlike(
+   $output,
+   qr/^nostalk-$/m,
+   "Retention test 2: tests, matched prefix-, are removed"
+);
+
+unlike(
+   $output,
+   qr/^nostalk-innodbstatus1$/m,
+   "Retention test 2: tests, matched prefix-innodbstatus1, are removed"
+);
+
+# ###########################################################################
+# Test if retention removes old files that match auto-generated pattern
+# ###########################################################################
+
+cleanup();
+
+system("mkdir $dest");
+$odate=`date --rfc-3339=date --date='-3 month'`;
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = `ls -l $dest | wc -l`;
+
+system("touch -d '$odate' $dest/*");
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = `ls -l $dest | wc -l` - $output;
+
+is(
+   $output,
+   0,
+   "Retention test 3: tests, matched auto-generated patern, are removed"
+);
+
+# ###########################################################################
+# Test if retention by size works as expected
+# ###########################################################################
+
+cleanup();
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --sleep 2 --dest $dest --pid $pid_file --iterations 5 -- --defaults-file=$cnf >$log_file 2>&1");
+
+$output = `du -s $dest | cut -f 1`;
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --retention-size 1 --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = $output / `du -s $dest | cut -f 1`;
+
+ok(
+   $output >= 5,
+   "Retention test 4: retention by size works as expected"
+);
+
+# ###########################################################################
+# Test if retention by count works as expected
+# ###########################################################################
+
+cleanup();
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --sleep 2 --dest $dest --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = `du -s $dest | wc -l`;
+
+$retval = system("$trunk/bin/pt-stalk --no-stalk --run-time 2 --dest $dest --retention-count 1 --pid $pid_file --iterations 1 -- --defaults-file=$cnf >$log_file 2>&1");
+
+PerconaTest::wait_until(sub { !-f $pid_file });
+
+$output = $output - `du -s $dest | wc -l`;
+
+is(
+   $output,
+   0,
+   "Retention test 5: retention by count works as expected"
+);
+
 # ###########################################################################
 # Test report about performance schema transactions in MySQL 5.7+
 # ###########################################################################
@@ -575,6 +708,7 @@ SKIP: {
    );
 }
 
+
 # #############################################################################
 # Done.
 # #############################################################################