issues
search
iati-data-access
/
data-backend
GNU Affero General Public License v3.0
1
stars
0
forks
source link
Review performance of API endpoints and model
#5
Open
markbrough
opened
1 year ago
markbrough
commented
1 year ago
Consider:
add additional indexes
further optimisation of queries (probably would need to be in
babbage
)
postgres version upgrade (e.g.
moving to v15 could involve a 10% improvement
)
psycopg2
to
psycopg
version upgrade (which is confusingly actually psycopg3) -
could be 4x-5x more memory efficient than
psycopg2
any optimisations that could be made in
iatidatacube/xlsx_writer.py
or whether it's necessary to move to e.g. a
redis
queue to generate the Excel files (would be nice to avoid that)
changing from
python
to
cython
?
Consider:
babbage)psycopg2topsycopgversion upgrade (which is confusingly actually psycopg3) - could be 4x-5x more memory efficient thanpsycopg2redisqueue to generate the Excel files (would be nice to avoid that)pythontocython?