On mg-rast in /etc/httpd.mg-rast-test/conf.d/meta.conf you will find: Alias /meta /vol/mg-rast-test/meta <Location /meta> AddHandler perl-script .cgi PerlResponseHandler ModPerl::Registry PerlOptions +ParseHeaders Options +ExecCGI Order allow,deny Allow from all </Location> Thus the meta server is /vol/mg-rast-test/meta/genome-meta-server.cgi it uses postgres database genome_meta on bio-big. Table meta_stash defines the storage for a meta file: genome_meta=# select * from meta_stash where meta_path = '/vol/mg-rast- prod/jobs/901/meta.xml'; id | creation_date | meta_path | server | job_id ----+----------------------------+------------------------------------- +--------+-------- 10 | 2008-06-23 15:59:30.657418 | /vol/mg-rast-prod/jobs/901/ meta.xml | | genome_meta=# select ts, key, value from meta_data where stash_id = 10 and key like 'status.%'; ts | key | value ----------------------------+------------------------ +--------------------------------------- 2008-06-23 16:48:01.993116 | status.check_sims | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 16:55:21.990019 | status.create_seed_org | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 16:57:37.264052 | status.export | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 17:02:26.958227 | status.final | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 16:16:32.94411 | status.preprocess | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 16:17:08.785478 | status.sims | <scalar><! [CDATA[complete]]></scalar> 2008-06-23 15:59:30.744101 | status.uploaded | <scalar><! [CDATA[complete]]></scalar> You can solve the copying-jobs problem by either changing the meta_path in the meta_stash table, or I suspect by inserting another row (nope, you can't; meta_stash.id is a primary key). So what your copy-job script could do is just update the meta_stash table. --bob