User 8a7878ec6d
16-04-2016 15:51:34
Hi,
I am importing an SD file (a vendor catalogue). I have tried to read it into a mySQL database, which caused the server to crash. Importing into a local Derby database is also extremely slow, already after the first 100 records it starts to slow down to parsing about one record per second. This is in IJC version 16.3.28.
I can provide you with the SD file if you so wish,maybe you can find something specific about it that causes the problem.
Best/Evert
ChemAxon 206bfdcce5
18-04-2016 08:35:54
Dear Evert,
we are not aware of any import related issues in recent versions of IJC and I was not able to reproduce this problem using our testing data (import on local MySQL - 1000 entries in 13s). Would you be so kind to provide us the problematic sdf file?
Best regards,
Karla
User 8a7878ec6d
18-04-2016 08:38:13
Do you have a place where I can upload the SD file in question?
Best/Evert
ChemAxon 206bfdcce5
02-05-2016 14:42:58
Dear Evert,
unfortunately I have no such place at this moment. Would it be possible for you to use some file sharing service and send to us the link? Or is the vendor catalog available online?
Best regards,
Karla
User 8a7878ec6d
03-05-2016 07:05:07
Hi,
Can you send me an e-mail address, so that I can share the SD file with you?
Thanks/Evert
ChemAxon 206bfdcce5
03-05-2016 07:11:39
Dear Evert,
could you please send the link to our support email ijc-support(at)chemaxon.com?
Thank you,
Karla
User 8a7878ec6d
03-05-2016 07:45:38
OK, I uploaded the SD file and shared it with the e-mail address you provided.
Best/Evert
ChemAxon 206bfdcce5
03-05-2016 07:58:51
Thank you, I will look into it.
Best regards,
Karla
ChemAxon 206bfdcce5
03-05-2016 09:46:35
Dear Evert,
I have tried to import the file both into Derby and MySQL (on localhost) and in my environment I did not encounter any issues as you describe - some data could not be imported (empty structures or similar problems, max 14 errors out of 100k+ structures), but in general it performed very well both in 'Molecules' and 'Any structures' table type, with MySQL (~1500s) being slower than Derby (~500s).
Have you encountered this issue with other sdf files of similar size (~100k records)?
Might it be possible that you run out of memory during import? Could you try to increase the Java heap size?
Best regards,
Karla