Closed carrickr closed 7 years ago
please don't forget the missing tiffs that need to be found https://github.com/nulib/repodev_planning_and_docs/issues/18
@Nic4Images
@jenyoung For display purposes we'll just use the jp2 display copies to make our display tiffs. This issue is dealing with improving performance.
We've moved all .jp2s that don't have a solr record to orphans
.
@davidschober We need a final call on what to do with the .jp2s
What's our confidence level that these were deleted from images (e.g. there's no associated metadata)?
Most of the orphans are likely from either deleted image records, batch testing or batch retries in sidekiq ( which used to create many orphaned derivatives during retry cycles but is now fixed)
Thanks. I'll dig into this a bit with the stakeholders. let's get a final size count on data when we're done. It sounds like our orphans were due to the system.
If the metadata is not there for these then it sounds okay to get rid of them.
i agree with Nicole. No metadata, no problem deleting
k. Let's get a count when we're done.
Cool. Let’s hold these for now.
On May 1, 2017, at 5:29 PM, Carrick Rogers notifications@github.com wrote:
Those without a record are:
oprhans.txt https://github.com/nulib/images/files/968934/oprhans.txt Counnt: 38759
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/nulib/images/issues/389#issuecomment-298449858, or mute the thread https://github.com/notifications/unsubscribe-auth/AElKntOlmFyNpMQhciyPyIYGcfdqFF39ks5r1lzQgaJpZM4NLkLn.
Any jp2 that doesn't exist in solr as a multires image does not need to be converted to .tiff. Find a location to move these rather than convert them.