Difference between revisions of "Brute force - modencode migration"
From WormBaseWiki
Jump to navigationJump to search(11 intermediate revisions by 2 users not shown) | |||
Line 5: | Line 5: | ||
* look at worm.conf | * look at worm.conf | ||
− | + | = Copy Databases = | |
+ | all the databases from modEncode that need to be moved over: henikoff hillier_genelets hillier_itranscripts lieb piano waterston snyder | ||
− | + | == Using mysqldump == | |
+ | The following instructions seem to cause errors with conflicting Storage versions in production (wb-www1) | ||
* dump databases using mysqldump (to find out which ones to get, look at conf files) | * dump databases using mysqldump (to find out which ones to get, look at conf files) | ||
− | + | mysqldump -uviewer -pviewer database_name > database_dump.sql | |
* copy database_dump.sql files to wb-dev | * copy database_dump.sql files to wb-dev | ||
* import databases | * import databases | ||
** first create database database_name | ** first create database database_name | ||
− | + | $ mysql -uroot -p | |
− | + | > CREATE DATABASE database_name; | |
+ | > exit | ||
+ | $ mysql -uroot -p database_name < database_dump.sql | ||
** grant select privileges to 'nobody' on the new database | ** grant select privileges to 'nobody' on the new database | ||
− | + | $ mysql -uroot -p | |
+ | > GRANT SELECT ON `database_name`.* TO 'nobody'@'localhost'; | ||
+ | > exit | ||
− | + | == Recreate Databases from GFF files == | |
− | * Write up how to do get GFF files and load | + | maybe this way will stop the Storage version problem? |
+ | * Write up how to do get GFF files and load to db - (bp_bulk_loader) | ||
+ | * GFF files located: gff/modencode/pi ... /gff3 | ||
+ | |||
+ | = Browser Data = | ||
+ | |||
+ | * copy relevant fasta file! this is used by some sam_bam tracks eg: piano_conf/piano_bam_miRNA_454.conf | ||
+ | /browser_data/worm/fasta/elegans.WS190.dna.fa & /browser_data/worm/fasta/elegans.WS190.dna.fa.fai | ||
− | |||
* copy relevant browser_data (wiggle files/sambam) to the same location on wb-dev | * copy relevant browser_data (wiggle files/sambam) to the same location on wb-dev | ||
** /browser_data/worm/wiggle_binaries/PI_NAME | ** /browser_data/worm/wiggle_binaries/PI_NAME | ||
** /browser_data/worm/sam_binaries/PI_NAME | ** /browser_data/worm/sam_binaries/PI_NAME | ||
+ | These are sparse files and should be copied compressed | ||
− | + | pack: | |
+ | tar -pczf ARCHIVE_NAME.tar.gz /browser_data/worm/wiggle_binaries/PI/ | ||
+ | unpack: | ||
+ | tar -xvzf ARCHIVE_NAME.tar.gz | ||
+ | |||
+ | = Config update = | ||
* copy relevant conf files to wb-dev | * copy relevant conf files to wb-dev | ||
** -> modencode: /var/www/conf/FOLDER_SPECIFIED_IN_WORM.CONF | ** -> modencode: /var/www/conf/FOLDER_SPECIFIED_IN_WORM.CONF | ||
** -> wb-dev : /usr/local/wormbase/gbrowse_current/conf/FOLDER_SPECIFIED_IN_WORM.CONF | ** -> wb-dev : /usr/local/wormbase/gbrowse_current/conf/FOLDER_SPECIFIED_IN_WORM.CONF | ||
* look at worm.conf and copy the appropriate #include function and database definitions to c_elegans.conf | * look at worm.conf and copy the appropriate #include function and database definitions to c_elegans.conf | ||
− | * in production (wb-www1), set remote renderer = | + | * in production (wb-www1), set remote renderer = 'none' for all the [TRACKS] |
+ | |||
+ | |||
+ | [[Category:Developer documentation]] |
Latest revision as of 22:56, 10 August 2010
This method does not lift-over the data to the current version of WormBase. This data will be mapped to WS190. This a temporary fix while we wait for the data to get lifted over at Sanger.
Get account on modencode.oicr.on.ca (email helpdesk@oicr.on.ca with request) this is their gbrowse production server.
- Configuration files: /var/www/conf/
- look at worm.conf
Contents
Copy Databases
all the databases from modEncode that need to be moved over: henikoff hillier_genelets hillier_itranscripts lieb piano waterston snyder
Using mysqldump
The following instructions seem to cause errors with conflicting Storage versions in production (wb-www1)
- dump databases using mysqldump (to find out which ones to get, look at conf files)
mysqldump -uviewer -pviewer database_name > database_dump.sql
- copy database_dump.sql files to wb-dev
- import databases
- first create database database_name
$ mysql -uroot -p > CREATE DATABASE database_name; > exit $ mysql -uroot -p database_name < database_dump.sql
- grant select privileges to 'nobody' on the new database
$ mysql -uroot -p > GRANT SELECT ON `database_name`.* TO 'nobody'@'localhost'; > exit
Recreate Databases from GFF files
maybe this way will stop the Storage version problem?
- Write up how to do get GFF files and load to db - (bp_bulk_loader)
- GFF files located: gff/modencode/pi ... /gff3
Browser Data
- copy relevant fasta file! this is used by some sam_bam tracks eg: piano_conf/piano_bam_miRNA_454.conf
/browser_data/worm/fasta/elegans.WS190.dna.fa & /browser_data/worm/fasta/elegans.WS190.dna.fa.fai
- copy relevant browser_data (wiggle files/sambam) to the same location on wb-dev
- /browser_data/worm/wiggle_binaries/PI_NAME
- /browser_data/worm/sam_binaries/PI_NAME
These are sparse files and should be copied compressed
pack:
tar -pczf ARCHIVE_NAME.tar.gz /browser_data/worm/wiggle_binaries/PI/
unpack:
tar -xvzf ARCHIVE_NAME.tar.gz
Config update
- copy relevant conf files to wb-dev
- -> modencode: /var/www/conf/FOLDER_SPECIFIED_IN_WORM.CONF
- -> wb-dev : /usr/local/wormbase/gbrowse_current/conf/FOLDER_SPECIFIED_IN_WORM.CONF
- look at worm.conf and copy the appropriate #include function and database definitions to c_elegans.conf
- in production (wb-www1), set remote renderer = 'none' for all the [TRACKS]