Using these commands can DESTROY your database if you don't know what your doing. I bear no responsiblities for your actions with these scripts. Use at your own risk.
Allows you to modify the data in system tables.
Cleans out catalog bloat from creating and deleting multiple objects. Run only during maintenence windows as it holds exclusive locks on many important system tables. Written with Lubomir Petrov
Print out disk space usage on DCA's. Can be helpful in identifying skew.
Collect the logs from a failed segment. Script needs to be updated with date of failure.
Shows where the segement's data directories are mapped to. Useful for identifying database data directories
Get the backtrace and query from a Greenplum core dump. Answers the question "what query caused the database to panic?"
Get the backtrace of all the core files that were generated on a DCA.
Run gpecheckcat on all databases and log it to a file. Run on a system with NO activity
gpcrondump used to hang. This script will cancel the gpcrondump in the right order using the right escalation of kill signals.
Extract the data from a gpexplorer output. Archaic.
Answers the question "What should I set my max_fsm to?"
Monitor active queries
gpssh without having to use a hostfile
GPDB used to have an issue where processes would get orphaned. It's been fixed, but this script will identify process that are no longer being tracked by GPDB (orphaned)
gdb scripts to work with gpdb cores. Written by Dan Lynch
postgres.gdb usage written by Dan Lynch
run gpcheckcat on all database, versions 4.1 and lower.
Get the backtrace and query from a Greenplum core dump. Answers the question "what query caused the database to panic?"
Redistribute a table based off of the OID. run this command offline. Written by Lubomir Petrov
Map data files to database id's.
Finds tables with OLD transcaction id's. Prevents Transaction ID wraparound (data loss) This corrects lack of maintenance.
Scripts specific to DCA hardware (mostly v1)
Does stuff with DCA v2 Super caps. Checks health, etc.. I didn't write this.
Check the health of the RAID controller batteries on a DCA v1.
Clear out the omreport logs when they fill up.
Cleanly shutdown a DCA.
DCA V1's could have performance issues with batteries on the RAID controllers, this forced this. Risk of dataloss for performance improvments...
Backup the switch config. Writen by Sagy Volkov