Parameters for mz_genenic runner
The process flow is controlled by setting few variables.
The most accurate and uptodate list of them is in the conf/_mz.conf
file.
There are few ways to specify these parameters.
- By using shell environment variables
- By placing (and editing)
_mz.conf
to the directorymz_generic.sh
is run from, - By specifying/setting a single shell variable
MZ_CONFIG=<path_to_conf_file>
with the patch to config file
Parameter | Example | Description | Comment |
---|---|---|---|
MZ_CONFIG |
/path/to/_mz.conf | Paths to the config files with the listed below options, sourced by mz_generic.sh |
|
MZ_RELEASE |
51 | EnsemblGenomes release version | |
ENS_VERSION |
104 | Ensembl (core db schema) release version | |
CMD |
cmd_server_alias | SQL DB server alias, to create core DB at and put e-hive DBs for running pipelines | |
PROD_SERVER |
prod_server_alias | Production SQL DB server alias to get ensembl_production and ncbi_taxonomy databases from |
Should be exported within config file (export PROD_SERVER ) |
For most of the use cases the meta configuration file (meta-file) should be provide as the first option:
./ensembl-production-metazoa/scripts/mz_generic.sh ${METACONF_DIR}/<meta_conf_file>
You can use either:
- abs-path
- or relative to the dir you're running scripts fromi
- or just name of the meta configuration file to be searched for in
./ensembl-production-metazoa/meta/${ENS_VERSION}
dir
I.e.
./ensembl-production-metazoa/scripts/mz_generic.sh ./ensembl-production-metazoa/meta/109_for_110/dmel
Though, you can use "help" and "env_setup_only" (see below) options instead of meta-file path on their own
./ensembl-production-metazoa/scripts/mz_generic.sh help
In addition to the meta conf paramete you can provide additional option, that controls the generic loader scripts/mz_generic.sh
behaviour:
Option | Description | Comments |
---|---|---|
help |
print usage and exit | |
env_setup_only |
prepare environment and exit | |
restore [tag_pattern] |
restore from the latest back up, or from the one matching "tag_pattern" | move bup symlink, if you need something older than the last saved; when pattern is used, back-ups that followed the macthing are removed (drops everything fresher then the one it's restoring from). N.B. as many ${DATA_DIR}/<meta_name>/done tags as you need (use ls -lt to sort tags by time). |
stop_after_conf |
get data, validate, prepare cofiguration for the loader and stop | |
stop_after_load |
stop after the loader pipeline before running anything else | |
stop_before_xref |
stop before running xref helper | |
pre_final_dc |
run datachecks before patching to the new schema stage | active, only if the goal is reachable |
finalise |
create back up with the final tag |
active, only if the goal is reachable |
patch_schema |
patch schema to the latest available |
Sometimes there's a need to have Ensembl environment, that can be used by unrelated pipelines / workflows.
To setup environment
configs_dir=$(pwd)
cd $configs_dir
git clone git@github.com:Ensembl/ensembl-production-metazoa.git
# source ensembl-production-metazoa/conf/_mz.conf file
# or
export PROD_SERVER=<prod_server_alias>
export ENS_VERSION=107
# setup environment (you should have proper github keys/permissions in place)
./ensembl-production-metazoa/scripts/mz_generic.sh env_setup_only 2>&1 | tee env_setup.log
cd ${configs_dir}
ln -s ensembl.prod.${ENS_VERSION} ensembl.prod
# source environment settings
source ${configs_dir}/ensembl.prod.${ENS_VERSION}/setup.sh