archiver/retriever: Handle USR1 signal to skip delay timers

Implement USR1 signal handlers to allow bypassing current delay timers
to force operations to proceed.

This is useful when doing manual testing or error recovery allowing
admins to skip the wait involved with ENDIT trying to perform tape-optimal
operations.

Solves #23

README.md

@@ -207,6 +207,19 @@ To run multiple instances for different tape pools on one host, the `ENDIT_CONFI
 to use a different configuration file. This is not to be confused with enabling parallel/multiple archive and
 retrieve operations for one pool which is done using options in the ENDIT daemon configuration file.
 
+# Bypassing delays/threshold/timers when testing
+
+The ENDIT daemons are designed to avoid unnecessary tape mounts, and achieves
+this by employing various thresholds and timers as explained in the example
+configuration file.
+
+However, when doing functional tests or error recovery related to the tape
+system it can be really frustrating having to wait longer than
+necessary. For these situations it's suitable to use the `USR1` signal
+handling in the ENDIT daemons. In general, the `USR1` signal tells the
+daemons to disregard all timers and thresholds and perform any pending
+actions immediately.
+
 # Collaboration
 
 It's all healthy perl, no icky surprises, we hope. Patches, suggestions, etc are

tsmarchiver.pl

@@ -34,6 +34,7 @@
 $Endit::logsuffix = 'tsmarchiver.log';
 my $filelist = "tsm-archive-files.XXXXXX";
 my $dsmcpid;
+my $skipdelays = 0; # Set by USR1 signal handler
 
 ##################
 # Helper functions
@@ -109,6 +110,7 @@ INIT
 $SIG{QUIT} = sub { warn("Got SIGQUIT, exiting...\n"); killchild(); exit; };
 $SIG{TERM} = sub { warn("Got SIGTERM, exiting...\n"); killchild(); exit; };
 $SIG{HUP} = sub { warn("Got SIGHUP, exiting...\n"); killchild(); exit; };
+$SIG{USR1} = sub { $skipdelays = 1; };
 
 my $desclong="";
 if($conf{'desc-long'}) {
@@ -125,6 +127,7 @@ INIT
 	getdir($dir, \%files);
 
 	if(!%files) {
+		$skipdelays = 0; # Ignore irrelevant request by USR1 signal
 		printlog "No files, sleeping for $conf{sleeptime} seconds" if($conf{debug});
 		sleep($conf{sleeptime});
 		next;
@@ -152,7 +155,11 @@ INIT
 		if(!defined($timer)) {
 			$timer = 0;
 		}
-		if($timer < $conf{archiver_timeout}) {
+		if($skipdelays) {
+			$skipdelays = 0; # Reset state set by USR1 signal
+			printlog "$usagestr below threshold and only waited $timer seconds, but proceeding anyway as instructed by USR1 signal";
+		}
+		elsif($timer < $conf{archiver_timeout}) {
 			if($conf{debug} || $conf{verbose} && $usagestr ne $lastusagestr) {
 				printlog "$usagestr below threshold, waiting for more data (waited $timer seconds)";
 			}

tsmretriever.pl

@@ -26,9 +26,13 @@
 
 # Add directory of script to module search path
 use lib dirname (__FILE__);
+
 use Endit qw(%conf readconf printlog);
 
+###########
+# Variables
 $Endit::logsuffix = 'tsmretriever.log';
+my $skipdelays = 0; # Set by USR1 signal handler
 
 # Turn off output buffering
 $| = 1;
@@ -60,6 +64,7 @@ ()
 $SIG{QUIT} = sub { warn("Got SIGQUIT, exiting...\n"); killchildren(); exit; };
 $SIG{TERM} = sub { warn("Got SIGTERM, exiting...\n"); killchildren(); exit; };
 $SIG{HUP} = sub { warn("Got SIGHUP, exiting...\n"); killchildren(); exit; };
+$SIG{USR1} = sub { $skipdelays = 1; };
 
 sub checkrequest($) {
 	my $req = shift;
@@ -335,13 +340,25 @@ ()
 			}
 
 			if($tape ne 'default' && defined $lastmount{$tape} && $lastmount{$tape} > time - $conf{retriever_remountdelay}) {
-				printlog "Skipping volume $tape, last mounted at " . strftime("%Y-%m-%d %H:%M:%S",localtime($lastmount{$tape})) . " which is more recent than remountdelay $conf{retriever_remountdelay}s ago" if($conf{verbose});
-				next;
+				my $msg = "volume $tape, last mounted at " . strftime("%Y-%m-%d %H:%M:%S",localtime($lastmount{$tape})) . " which is more recent than remountdelay $conf{retriever_remountdelay}s ago";
+				if($skipdelays) {
+					printlog "Proceeding due to USR1 signal despite $msg";
+				}
+				else {
+					printlog "Skipping $msg" if($conf{verbose});
+					next;
+				}
 			}
 
 			if($tape ne 'default' && $job->{$tape}->{tsoldest} > time()-$conf{retriever_reqlistfillwaitmax} && $job->{$tape}->{tsnewest} > time()-$conf{retriever_reqlistfillwait}) {
-				printlog "Skipping volume $tape, request list $job->{$tape}->{listsize} entries and still filling, oldest " . strftime("%Y-%m-%d %H:%M:%S",localtime($job->{$tape}->{tsoldest})) . " newest " .  strftime("%Y-%m-%d %H:%M:%S",localtime($job->{$tape}->{tsnewest})) if($conf{verbose});
-				next;
+				my $msg = "volume $tape, request list $job->{$tape}->{listsize} entries and still filling, oldest " . strftime("%Y-%m-%d %H:%M:%S",localtime($job->{$tape}->{tsoldest})) . " newest " .  strftime("%Y-%m-%d %H:%M:%S",localtime($job->{$tape}->{tsnewest}));
+				if($skipdelays) {
+					printlog "Proceeding due to USR1 signal despite $msg";
+				}
+				else {
+					printlog "Skipping $msg" if($conf{verbose});
+					next;
+				}
 			}
 
 			my ($lf, $listfile) = tempfile("$tape.XXXXXX", DIR=>"$conf{dir}/requestlists", UNLINK=>0);
@@ -507,4 +524,5 @@ ()
 			}
 		}
 	}
+	$skipdelays = 0; # Reset state set by USR1 signal
 }