-
We have FMCG application in Iraq, we use ERPnext version-13 as the backend, but as we grow we face a lot of problems may be becuase Stander ERPnext can’t handle a lot of requests, becuase one of the common problems we face when running a huge report (10k row) the report, the system is down (Request time out) or (server is too busy)
-
I need someone who has good experience in handling those issues and advise us on what we use and how is use
How you can help me with this case
Thank you @Antony_Praveenkumar, can you explain the first point
- Use NumPy & Pandas package to handle arithmatic operations.
what do you mean about this? and where I can use it.
Hi @Omar_Mohammed
I have a customer that is a telecom company, and another a furniture industry!
Both have similar situations, telecom need to generate thousands of invoices in a short period of time, and furniture create hundreds of lines in BOMs due product complexity!
What I learned:
- Database Indexes boost the performance a lot
- Don’t use the same disk for file I/O, Db I/O and Proc I/O
- Cachê everything, caching functions like memoize and LRU are your best friends, also use redis or memcached where you can
- if you need to deal with large amount of data in memory, try to use lazy methods, coroutines and generators (yield functions in python)
- Finally, split the work where you can in different background works, to be processed as batches, this will boost your performance!
4 Likes
NumPy & Pandas as are very important package in python its 50x Faster than normal Python List
if you want to make any custom report
I’m recommended to use these packages.
and one more better options is move your Versio -14 to Version-15 because in Version-15 they have improved 800+ place regarding the performance. its one more better options
1 Like