Opened 5 years ago
Closed 4 years ago
#2264 closed task (worksforme)
make sure rasdaman with postgres compiles and passes systemtest
Reported by: | Dimitar Misev | Owned by: | apercov |
---|---|---|---|
Priority: | major | Milestone: | 10.0 |
Component: | reladminif | Version: | 9.8 |
Keywords: | Cc: | ||
Complexity: | Medium |
Description
- Configure rasdaman with postgresql backend (instead of sqlite)
cmake ... -DDEFAULT_BASEDB=postgresql
- Build with make/make install
- it may require fixing compilation in *.pgc files in rel* source directories
- especially with axis names, see changeset:7a60837d3bb57f799abb064e0470d0e39afba2f5 which only updates the sqlite *.sc files and not *.pgc
- Initialize db with create_db.sh and update_db.sh
- check this first
- Run systemtest and fix possible issues
Change History (5)
comment:1 by , 5 years ago
Status: | assigned → accepted |
---|
comment:2 by , 5 years ago
comment:4 by , 5 years ago
Important notice: Test_manipulation fails every time in RELEASE on my local machine. Few times it corrupted the db, other times it went to deadlock state (my assumption). To see that the transaction gets stuck, you can open psql command line, connect to RASBASE(or any other name of base DB that you are using), and run the following query:
SELECT * FROM pg_stat_activity WHERE datname = current_database() AND pid <> pg_backend_pid() AND state LIKE 'idle%';
Version 0, edited 5 years ago by (next)
comment:5 by , 4 years ago
Resolution: | → worksforme |
---|---|
Status: | accepted → closed |
Note:
See TracTickets
for help on using tickets.
Both Rasdapy and Rasdapy3 work unreliably with postgres. Rasdapy3 cannot pass the testcases due to that. Should I continue work on that in this ticket, or should I open another ticket for that and just disable Rasdapy3 test for postgres?