#community-help

Issues with Docker Image Data Upload

TLDR Nelson encountered errors while uploading data using the-typesense/typesense:0.22.0.rcs16 image. Kishore Nallan suggested using typesense/typesense:0.22.0.rcs17 instead, but the issue persisted.

Powered by Struct AI
eyes1
Oct 16, 2021 (23 months ago)
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
07:56 AM
Hi Jason Kishore Nallan I have this mistake when you upload the data in lots
07:58
Nelson
07:58 AM
Docker image typesense/typesense:0.22.0.rcs16
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:02 AM
I saw the issue you created on GitHub and replied there: https://github.com/typesense/typesense/issues/403#issuecomment-944252233
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
08:08 AM
ok
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:12 AM
You need to use rcs17 and not rcs16
eyes1
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
08:17 AM
already change version, I'm climbing the data again to see if it produces the same error
08:31
Nelson
08:31 AM
Error
08:31
Nelson
08:31 AM
Jason Kishore Nallan
08:31
Nelson
08:31 AM
typesense/typesense:0.22.0.rcs17
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:35 AM
How long does it take to trigger this issue?
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
08:40 AM
I did everything from 0 in a single node. I use the Upload Data script of Python SDK. We had approximately this amount of records uploaded
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:43 AM
I believe this might be a regression on the recent 0.22.rcs builds. I've DMed you on a course of action.
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
08:46 AM
file
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:47 AM
How many records are there totally? If it fails within the first 1M records then might not need the full file to trigger the issue.
Nelson
Photo of md5-04a94d0e1b53117c272637eb497a1540
Nelson
08:51 AM
11,094,533. lines

I have it segmented by pieces of 1,000,000 records too
08:52
Nelson
08:52 AM
Do you think I do if I do it by segments of 1,000,000 records do not fail ??
Kishore Nallan
Photo of md5-4e872368b2b2668460205b409e95c2ea
Kishore Nallan
08:53 AM
If it's possible for you to share the 1M dataset with me (you can upload to Google Drive and share the link) then I can try to reproduce the issue. Or we need a coredump of the crash to analyze what's happening.

I am not able to trigger this issue on the large datasets I have tested with.