diff --git a/Dockerfile b/Dockerfile
index a4fcc566..5177de19 100755
--- a/Dockerfile
+++ b/Dockerfile
@@ -45,7 +45,7 @@ RUN apt-get install -y python-dev python-pip && \
   pip install \
   openpyxl \
   gspread \
-  xlrd \
+  xlrd==1.2.0 \
   lepl \
   requests_toolbelt \
   cassandra-driver \
diff --git a/src/MGRAST/bin/validate_metadata b/src/MGRAST/bin/validate_metadata
index 90276a88..0a759d6c 100755
--- a/src/MGRAST/bin/validate_metadata
+++ b/src/MGRAST/bin/validate_metadata
@@ -285,8 +285,8 @@ def google_to_mdraw(url, dupe, inc, login, password):
       return tables_to_mdraw(tables, dupe)
 
 def xlsx_to_mdraw(xxfile, dupe, inc):
-    xxbook = load_workbook(filename = xxfile)
-    tables = dict([(x, sheet_to_table(xxbook[x], 'xlsx')) for x in xxbook.sheetnames()])
+    xxbook = load_workbook(filename = xxfile, data_only="True")
+    tables = dict([(x, sheet_to_table(xxbook[x], 'xlsx')) for x in xxbook.sheetnames])
     if inc:
       return tables_to_mdraw_inc(tables, dupe)
     else:
@@ -307,9 +307,18 @@ def sheet_to_table(sheet, source):
         maxCol = sheet.max_column
         for r in range(maxRow):
             row = []
-            for c in range(maxCol):
-                row.append( sheet.cell(row=r,column=c).value )
-            table.append(row)
+            rowcontents =  [ sheet.cell(row=r+1, column=i+1).value for i in range(maxCol)]
+            # do not add empty rows, unless there is only one column
+            if not all([ element is None for element in rowcontents] ) or maxCol == 1 :  
+                for c in range(maxCol):
+                    value = sheet.cell(row=r+1,column=c+1).value 
+                    # sanitize None values to empty strings 
+                    if value is None:
+                       row.append("")
+                    else:
+                       row.append( value )
+                table.append(row)
+
     if source == 'xls':
         maxRow = sheet.nrows
         maxCol = sheet.ncols
@@ -691,7 +700,7 @@ def main(args):
             try:
                 mdraw = xls_to_mdraw(md_source, opts.dupe, opts.inc)
             except:
-                errorMsg('Invald file format: '+md_source, 1)
+                errorMsg('Error loading xlsx, xls file: '+md_source, 1)
     elif opts.format == 'xls':
         try:
             mdraw = xls_to_mdraw(md_source, opts.dupe, opts.inc)
@@ -699,7 +708,7 @@ def main(args):
             try:
                 mdraw = xlsx_to_mdraw(md_source, opts.dupe, opts.inc)
             except:
-                errorMsg('Invald file format: '+md_source, 1)
+                errorMsg('Error loading xls, xlsx file: '+md_source, 1)
     elif opts.format == 'google':
         mdraw = google_to_mdraw(md_source, opts.dupe, opts.inc, opts.login, opts.password)
     #pprint.pprint(mdraw)
diff --git a/src/MGRAST/lib/Auth.pm b/src/MGRAST/lib/Auth.pm
index af20150f..0387e69d 100644
--- a/src/MGRAST/lib/Auth.pm
+++ b/src/MGRAST/lib/Auth.pm
@@ -89,7 +89,10 @@ sub authenticate {
 	    } elsif (exists($response->{error}) && $response->{error}) {
 	      $verbose.=', "preferences": "ERROR - '.$response->{error}.'"';
 	    } else {
+		 $verbose.=', "preferences": "" ';
+		eval{  # Users with corrupt preferences were being denied access here (mgu21652)
 	      $verbose.=', "preferences": '.$json->encode($response->{data}->{attributes}->{pref});
+		}
 	    }
 	  }
 	  if ($us->has_right(undef, 'edit', 'user', '*')) {
diff --git a/src/MGRAST/lib/ElasticSearch.pm b/src/MGRAST/lib/ElasticSearch.pm
index f496e684..92bc6877 100644
--- a/src/MGRAST/lib/ElasticSearch.pm
+++ b/src/MGRAST/lib/ElasticSearch.pm
@@ -129,6 +129,8 @@ our $fields = {
 	filter_ambig => 'pipeline_parameters_filter_ambig',
 	filter_ln => 'pipeline_parameters_filter_ln',
 	bowtie => 'pipeline_parameters_bowtie',
+	project_tags => 'project_tags',
+	sample_tags => 'sample_tags',
 };
 
 our $prefixes = {
@@ -163,6 +165,7 @@ our $prefixes = {
 		'project_funding',
 		'project_name',
 		'project_id',
+		'tags',
 	],
 	'sample_' => [
 		'biome_id',
@@ -191,6 +194,7 @@ our $prefixes = {
 		'feature',
 		'metagenome_taxonomy_id',
 		'env_package',
+		'tags',
 	],
 	'env_package_' => [
 		'env_package_id',
@@ -321,6 +325,8 @@ our $types = {
 	filter_ambig => 'boolean',
 	filter_ln => 'boolean',
 	bowtie => 'boolean',
+	project_tags => 'arrayText',
+	sample_tags => 'arrayText',
 };
 
 our $taxa_num = [
diff --git a/src/MGRAST/lib/resources/heartbeat.pm b/src/MGRAST/lib/resources/heartbeat.pm
index 3b2a69fd..de3da6a1 100644
--- a/src/MGRAST/lib/resources/heartbeat.pm
+++ b/src/MGRAST/lib/resources/heartbeat.pm
@@ -18,7 +18,7 @@ sub new {
   $self->{name} = "heartbeat";
   $self->{m5nr_version} = 1;
   $self->{services} = {
-      'FTP' => 'ftp://'.$Conf::ftp_download,
+#      'FTP' => 'ftp://'.$Conf::ftp_download,
       'website' => $Conf::web_site,
       'SHOCK' => $Conf::shock_url,
       'SHOCKDB' => 'mongo',
@@ -30,7 +30,7 @@ sub new {
       'cassandra' => $Conf::cassandra_m5nr
   };
   $self->{attributes} = { "service" => [ 'string', "cv", [
-                               ['FTP', 'file server'],
+#                               ['FTP', 'file server'],
 							   ['website', 'MG-RAST website'],
 							   ['SHOCK', 'object storage'],
 							   ['SHOCKDB', 'object storage mongodb'],
diff --git a/src/MGRAST/lib/resources/resource.pm b/src/MGRAST/lib/resources/resource.pm
index 936466e5..3071695f 100644
--- a/src/MGRAST/lib/resources/resource.pm
+++ b/src/MGRAST/lib/resources/resource.pm
@@ -1369,7 +1369,7 @@ sub get_shock_preauth {
     my $response = undef;
     eval {
         my @args = $auth ? ('Authorization', "$authPrefix $auth") : ();
-        my $get = $self->agent->get($Conf::shock_url.'/node/'.$id.'?download_url'.($fn ? "&filename=".$fn : ""), @args);
+        my $get = $self->agent->get($Conf::shock_url.'/node/'.$id.'?download_url'.($fn ? "&file_name=".$fn : ""), @args);
         $response = $self->json->decode( $get->content );
     };
     if ($@ || (! ref($response))) {
@@ -2277,13 +2277,12 @@ sub upsert_to_elasticsearch_annotation {
     $self->json->utf8();
     my $success = {};
     
-    # PUT docuemnt(s)
+    # PUT document(s)
     foreach my $key (keys %$results) {
         if ($results->{$key}) {
             my $entry = $self->json->encode($results->{$key});
             my $esurl = $Conf::es_host."/$index/$key/$mgid?parent=$mgid";
             my $response = undef;
-            eval {
                 my @args = (
                     'Content_Type', 'application/json',
                     'Content', $entry
@@ -2291,10 +2290,12 @@ sub upsert_to_elasticsearch_annotation {
                 my $req = POST($esurl, @args);
                 $req->method('PUT');
                 my $put = $self->agent->request($req);
+                $response = $put->content;  # If it's an error message, preserve it.
+            eval {
                 $response = $self->json->decode($put->content);
             };
             if ($@ || (! ref($response)) || $response->{error} || (! $response->{result})) {
-                $success->{$key} = "failed";
+                $success->{$key} = "failed: $response";
             }
             $success->{$key} = "updated";
         } else {
@@ -2415,19 +2416,29 @@ sub get_elastic_query {
     }
     
     my $content;
+    my $res;  # need it here for error trapping
     eval {
-        my $res  = $self->agent->post($server.'/_search', Content => $self->json->encode($postJSON));
+        $res  = $self->agent->post($server.'/_search', Content => $self->json->encode($postJSON), 
+                                      "Content-Type", "application/json");
         $content = $self->json->decode($res->content);
     };
     if ($@ || (! ref($content))) {
+        # one of the likely things to get us here is that json->decode fails because error message, not json
+        $self->return_data( { "ERROR" => $res->content} , 500) ; 
         return undef, $@;
     } elsif (exists $content->{error}) {
-        if (exists($content->{error}{type}) && exists($content->{error}{reason}) && exists($content->{status})) {
-            $self->return_data( {"ERROR" => $content->{error}{type}.": ".$content->{error}{reason}}, $content->{status} );
-        } else {
-            $self->return_data( {"ERROR" => "Invalid Elastic Search return response"}, 500 );
-        }
-    } else {
+        if (ref $content->{error}) {
+         eval {
+              if (exists($content->{error}{type}) && exists($content->{error}{reason}) && exists($content->{status})) {
+                  $self->return_data( {"ERROR" => $content->{error}{type}.": ".$content->{error}{reason}.$res->$content}, $content->{status} );
+              } else {
+                 $self->return_data( {"ERROR" => "Invalid Elastic Search return response: $res"}, 500 );
+              }
+        # fallback -- if error, but eval aborts 
+        $self->return_data( { "ERROR" => $res->content }, 500) ; } 
+        } else {  # content->error is not ref
+        $self->return_data( { "ERROR" => "ES response: $content->{error}" }, 500) ; }
+    } else {   #no error
         return $content;
     }
 }
@@ -3438,6 +3449,10 @@ sub jsonTypecast {
         $val =~ s/^\s+//;
         $val =~ s/\s+$//;
         $val =~ s/\s+/ /g;
+    } elsif ( $type eq 'arrayText' ) {
+        my @array  = split(';', $val);
+        $val = \@array; 
+         
     } elsif (($type eq 'integer') || ($type eq 'long')) {
         if ($val =~ /^[+-]?\d+$/) {
             $val = int($val);
diff --git a/src/MGRAST/lib/resources/search.pm b/src/MGRAST/lib/resources/search.pm
index 80201128..817ff272 100644
--- a/src/MGRAST/lib/resources/search.pm
+++ b/src/MGRAST/lib/resources/search.pm
@@ -243,10 +243,18 @@ sub query {
             $query =~ s/^\s+|\s+$//g;
             $query =~ s/\s+/ /g;
             # remove specified fields (non-escaped ':'), only using set default
-            if (($query =~ /:/) && ($query !~ /\\:/)) {
+            # if (($query =~ /:/) && ($query !~ /\\:/)) {
+            #     my @parts = split(/:/, $query);
+            #     $query = join(" ", @parts[1..$#parts]);
+            # }
+            # remove specified fields (non-escaped ':'), only using set default
+            if (( $field !~ /_tag/ ) && ($query =~ /:/) && ($query !~ /\\:/)) {
                 my @parts = split(/:/, $query);
                 $query = join(" ", @parts[1..$#parts]);
             }
+            else {
+		        $query=~s/\:/\\\:/g
+            }
             push @$queries, {"field" => $key, "query" => $query, "type" => $type};
         }
     }
diff --git a/src/MGRAST/pylib/mgrast_cassandra.py b/src/MGRAST/pylib/mgrast_cassandra.py
index 3208c485..aa84cb10 100644
--- a/src/MGRAST/pylib/mgrast_cassandra.py
+++ b/src/MGRAST/pylib/mgrast_cassandra.py
@@ -246,7 +246,7 @@ def get_md5_records(self, job, swap=None, md5s=None, evalue=None, identity=None,
         rmqLogger(self.channel, 'select', query)
         rows = self.session.execute(query)
         for r in rows:
-            if r[1] == 0:
+            if r[1] == 0 or r[0] is None:  # skip row if zero length, or row is corrupt
                 continue
             pos = bisect.bisect(found, (r[0], None))
             if (pos > 0) and ((found[pos-1][0] + found[pos-1][1]) == r[0]):
diff --git a/src/PPO/PPOBackend.pm b/src/PPO/PPOBackend.pm
index e41b6789..c57cfeac 100644
--- a/src/PPO/PPOBackend.pm
+++ b/src/PPO/PPOBackend.pm
@@ -209,7 +209,7 @@ sub get_rows {
       ? ' DESC' : ' ASC';
   }
   
-  my $statement = sprintf ("SELECT %s FROM %s%s%s",
+  my $statement = sprintf ("SELECT %s FROM %s%s%s%s",
 			   (@$fields) ? join (",", @$fields) : '*',
 			   $table,
 			   ($conditions) ? " WHERE $conditions" : '',