Open ralwing opened 5 months ago
Try to set chunk size less than default (for example 16 images). One more suggestion is to disable "Use cache" option. It will make task creating longer.
I have other sets of pointclouds which are way bigger than this (10GB), and they are uploaded and opened fine.
Redis logs:
I20240529 07:42:07.455459 104 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 07:59:59.830586 103 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 08:12:58.690346 102 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 08:13:03.894992 105 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 08:28:58.435904 104 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 09:23:53.568164 104 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 09:24:00.185640 102 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
I20240529 09:43:35.570396 106 redis_connection.cc:88] [connection] Failed to tokenize the request. Error: Protocol error: invalid bulk length
Shrinking the buffer didn't help.
Although disabling no-cache
option helps, it triples the job size.
Actions before raising this issue
Steps to Reproduce
Create a new task with this collection of pcd files:
Try to open the job
The server returns 500 while trying to load and in there is a crash in the logs
Expected Behavior
The job is opening without a crashing server.
Possible Solution
When I create a task by splitting this set of files into several smaller ones, the task loads correctly
Context
No response
Environment