Misc. minor dev

modules/Bio/EnsEMBL/Hive/Worker.pm

@@ -471,7 +471,7 @@ sub toString {
             $include_analysis ? ( 'analysis='.($current_role ? $current_role->analysis->logic_name.'('.$current_role->analysis_id.')' : 'UNSPECIALIZED') ) : (),
             'resource_class_id='.($self->resource_class_id // 'NULL'),
             'meadow='.$self->meadow_type.'/'.$self->meadow_name,
-            'process='.$self->meadow_user.'@'.$self->meadow_host.'#'.$self->process_id,
+            'process='.$self->meadow_user.($self->meadow_host ? '@'.$self->meadow_host : 'UNALLOCATED').'#'.$self->process_id,
             'when_checked_in='.($self->when_checked_in // 'NEVER'),
             'batch_size='.($current_role ? $current_role->analysis->stats->get_or_estimate_batch_size() : 'UNSPECIALIZED'),
             'job_limit='.($self->job_limiter->available_capacity() // 'NONE'),

sql/procedures.mysql

@@ -32,7 +32,7 @@ CONTACT
 --
 -- Usage:
 --       select * from progress;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
---       select * from progress where logic_name like 'family_blast%';   # only show family_blast-related analyses
+--       select * from progress where analysis_name_and_id like 'family_blast%';   # only show family_blast-related analyses
 --       select * from progress where retry_count>1;                     # only show jobs that have been tried more than once
 
 CREATE OR REPLACE VIEW progress AS
@@ -49,6 +49,37 @@ CREATE OR REPLACE VIEW progress AS
     ORDER BY a.analysis_id, j.status;
 
 
+-- show the hierarchy of hive jobs (which analysis seeded jobs where) ----------------------------
+--
+-- Usage:
+--       select * from job_hierarchy;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
+--       select * from job_hierarchy where retry_count>1;                     # only show jobs that have been tried more than once
+
+CREATE OR REPLACE VIEW job_hierarchy AS
+	SELECT
+		CONCAT(a1.logic_name, '(', a1.analysis_id, ')') AS analysis_name_and_id,
+		CONCAT(a2.logic_name, '(', a2.analysis_id, ')') AS parent_analysis_name_and_id,
+		j1.status AS status,
+		j1.retry_count AS retry_count,
+		COUNT(0) AS cnt,
+		MIN(j1.job_id) AS example_job_id
+	FROM
+		job j1
+		LEFT JOIN job j2 ON (j1.prev_job_id = j2.job_id)
+		LEFT JOIN analysis_base a1 ON (j1.analysis_id = a1.analysis_id)
+		LEFT JOIN analysis_base a2 ON (j2.analysis_id = a2.analysis_id)
+	GROUP BY
+		a1.logic_name,
+		a2.logic_name,
+		j1.status,
+		j1.retry_count
+	ORDER BY
+		a1.analysis_id,
+		a2.analysis_id,
+		j1.retry_count;
+DROP VIEW job_hierarchy;
+
+
 -- a convenient view that also incorporates (otherwise redundant) analysis_id and logic_name ------------
 --
 -- Usage:

sql/procedures.pgsql

@@ -32,7 +32,7 @@ CONTACT
 --
 -- Usage:
 --       select * from progress;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
---       select * from progress where logic_name like 'family_blast%';   # only show family_blast-related analyses
+--       select * from progress where analysis_name_and_id like 'family_blast%';   # only show family_blast-related analyses
 --       select * from progress where retry_count>1;                     # only show jobs that have been tried more than once
 
 CREATE OR REPLACE VIEW progress AS
@@ -49,6 +49,37 @@ CREATE OR REPLACE VIEW progress AS
     ORDER BY a.analysis_id, j.status;
 
 
+-- show the hierarchy of hive jobs (which analysis seeded jobs where) ----------------------------
+--
+-- Usage:
+--       select * from job_hierarchy;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
+--       select * from job_hierarchy where retry_count>1;                     # only show jobs that have been tried more than once
+
+CREATE OR REPLACE VIEW job_hierarchy AS
+	SELECT
+		a1.logic_name || '(' || a1.analysis_id || ')' AS analysis_name_and_id,
+		a2.logic_name || '(' || a2.analysis_id || ')' AS parent_analysis_name_and_id,
+		j1.status AS status,
+		j1.retry_count AS retry_count,
+		COUNT(0) AS cnt,
+		MIN(j1.job_id) AS example_job_id
+	FROM
+		job j1
+		LEFT JOIN job j2 ON (j1.prev_job_id = j2.job_id)
+		LEFT JOIN analysis_base a1 ON (j1.analysis_id = a1.analysis_id)
+		LEFT JOIN analysis_base a2 ON (j2.analysis_id = a2.analysis_id)
+	GROUP BY
+		a1.logic_name,
+		a2.logic_name,
+		j1.status,
+		j1.retry_count
+	ORDER BY
+		a1.analysis_id,
+		a2.analysis_id,
+		j1.retry_count;
+DROP VIEW job_hierarchy;
+
+
 -- a convenient view that also incorporates (otherwise redundant) analysis_id and logic_name ------------
 --
 -- Usage:

sql/procedures.sqlite

@@ -32,7 +32,7 @@ CONTACT
 --
 -- Usage:
 --       select * from progress;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
---       select * from progress where logic_name like 'family_blast%';   # only show family_blast-related analyses
+--       select * from progress where analysis_name_and_id like 'family_blast%';   # only show family_blast-related analyses
 --       select * from progress where retry_count>1;                     # only show jobs that have been tried more than once
 
 DROP VIEW IF EXISTS progress;
@@ -50,6 +50,38 @@ CREATE VIEW IF NOT EXISTS progress AS
     ORDER BY a.analysis_id, j.status;
 
 
+-- show the hierarchy of hive jobs (which analysis seeded jobs where) ----------------------------
+--
+-- Usage:
+--       select * from job_hierarchy;                                         # the whole table (may take ages to generate, depending on the size of your pipeline)
+--       select * from job_hierarchy where retry_count>1;                     # only show jobs that have been tried more than once
+
+DROP VIEW IF EXISTS job_hierarchy;
+CREATE VIEW IF NOT EXISTS job_hierarchy AS
+	SELECT
+		a1.logic_name || '(' || a1.analysis_id || ')' AS analysis_name_and_id,
+		a2.logic_name || '(' || a2.analysis_id || ')' AS parent_analysis_name_and_id,
+		j1.status AS status,
+		j1.retry_count AS retry_count,
+		COUNT(0) AS cnt,
+		MIN(j1.job_id) AS example_job_id
+	FROM
+		job j1
+		LEFT JOIN job j2 ON (j1.prev_job_id = j2.job_id)
+		LEFT JOIN analysis_base a1 ON (j1.analysis_id = a1.analysis_id)
+		LEFT JOIN analysis_base a2 ON (j2.analysis_id = a2.analysis_id)
+	GROUP BY
+		a1.logic_name,
+		a2.logic_name,
+		j1.status,
+		j1.retry_count
+	ORDER BY
+		a1.analysis_id,
+		a2.analysis_id,
+		j1.retry_count;
+DROP VIEW job_hierarchy;
+
+
 -- a convenient view that also incorporates (otherwise redundant) analysis_id and logic_name ------------
 --
 -- Usage:

t/apache2.t → t/00.housekeeping/apache2.t

@@ -38,7 +38,7 @@ my $original_dir = cwd();
 chdir($file_dir);
 my $cur_dir = cwd();
 chdir($original_dir);
-my $root = File::Spec->catdir($cur_dir, File::Spec->updir());
+my $root = File::Spec->catdir($cur_dir, File::Spec->updir(), File::Spec->updir());
 
 
 =head2 has_apache2_licence

t/ascii.t → t/00.housekeeping/ascii.t

@@ -40,7 +40,7 @@ my $original_dir = cwd();
 chdir($file_dir);
 my $cur_dir = cwd();
 chdir($original_dir);
-my $root = File::Spec->catdir($cur_dir, File::Spec->updir());
+my $root = File::Spec->catdir($cur_dir, File::Spec->updir(), File::Spec->updir());
 
 
 sub is_ascii {

t/perlcritic.t → t/00.housekeeping/perlcritic.t

@@ -46,7 +46,7 @@ my $original_dir = cwd();
 chdir($file_dir);
 my $cur_dir = cwd();
 chdir($original_dir);
-my $root = File::Spec->catdir($cur_dir, File::Spec->updir());
+my $root = File::Spec->catdir($cur_dir, File::Spec->updir(), File::Spec->updir());
 
 # Configure critic
 Test::Perl::Critic->import(-profile => File::Spec->catfile($root, 'perlcriticrc'), -severity => 5, -verbose => 8);