24000 items performance


I’ve been filling a Zotero library from an off line bibliography using the API. The library contains 24000+ items. However the web client seems very sluggish with these numbers of items. Is there any way to make the web client more useable? Maybe by reducing the number of collections or tags? The library is located here: https://www.zotero.org/erichkaestner/library

Thanks for any comments / pointers.
  • edited September 18, 2021
    We can look into the performance, but is this for personal usage or for public consumption? For personal usage, you'd very much want to use the desktop app for a library of that size (and in general for any significant work with Zotero).
  • hey @erichkaestner, thanks for your feedback on performance, which part of the interface would you say feels sluggish? Is it the fact that you see placeholders while waiting for items to load? Also what browser are you using?
  • We've ported this bibliography to Zotero with the intention of making it publicly available. The issue is indeed that the item placeholders sometimes take a long time to load, or they don't seem to load at all (I have waited several minutes). This happens even on small subcollections. (Example: https://ibb.co/M877hhJ)
    The collections, tags, and "xxx items in this view" elements do always load fine.

    I have tried Edge, Chrome and Brave browsers and have the same issue in all of them.

    Don't know if it's relevant, but during the conversion to Zotero we have made several "attempts" which we then deleted, and re-uploaded with some adaptations.

    Thanks for looking in to this!
  • I have to say Zotero is performing a lot better today. I've had no more incidents were item placeholders are not loaded and the loading times seem fine. Did the system need time to build the indexes? Or is this the result of some tweaking on your end?

  • We did fix a couple issues that may have been impacting performance.

    Definitely let us know if this problem returns and we can look more into it.
  • Thank you very much!
Sign In or Register to comment.