search

sanger / limber

A config-driven LIMS built on Sequencescape, primarily for running library preparation pipelines in the laboratory
MIT License
4 stars 8 forks source link

Y24-303 - [BUG] Bioscan - Exception - A Net::ReadTimeout occurred in work_completions#create: #1899

Open andrewsparkes opened 2 months ago

andrewsparkes commented 2 months ago

Describe the bug User clicks the charge and pass button on the XP tube in Bioscan and gets a timeout exception. See Limber server exception at 14.33pm on 5th Sept 2024.

RT Ticket Number NA

To Reproduce Steps to reproduce the behaviour:

  1. User clicked on 'charge and pass' button on the LBSN-9216 Lib PCR Pool XP tube at the end of the Bioscan pipeline
  2. Because there are so many samples in this tube the system times out

Expected behaviour The 'charge and pass' button should not appear on this tube. Charge and pass happens on the PCR 2 plate earlier in this pipeline.

Screenshots NA

Desktop (please complete the following information): NA

Additional context Normally in a library prep pipeline we would charge and pass on the XP labware at the end to close off the library prep requests. Bioscan is a bit unusual in that we create 24 x PCR 2 plates (under different submissions) then pool those together to get to a single tube containing up to 9216 samples, and we found during development that we had to close off the library prep requests before pooling (restriction on pooling multiple submissions with active requests?). So the charge and pass is configured on the PCR 2 plate in the Limber pipeline config. However, for some reason, we see a charge and pass button on BOTH the PCR 2 plate and the XP tube labwares.

2 options:

  1. determine why a charge and pass button appears on the XP tube despite not being set on the pipeline config and prevent it showing (so they can't click it and trigger a timeout) - easiest option
  2. figure out how to allow the pooling without closing off the requests, and have the charge and pass on the XP tube to align with other pipelines - much harder option, would have to change the code around pooling to allow multiple submissions to be pooled, and in charge and passing to allow requests from multiple submissions to be closed, plus make the charge and passing on a tube with 9216 samples efficient enough that it doesn't trigger a time out exception (possibly needing an asynchronous solution). Also a change to lab staff training / SOP to charge and pass at the different point.
stevieing commented 1 month ago

There is a workaround so only happens if they do the wrong thing.

stevieing commented 1 month ago

Could wait until there is a refactor of Bioscan.