We're running a 8.17.2 fairly fresh umbraco site on a 2019 windows server with a MSSQL db. Noticed lately that certain documents, seemingly random ones, will take 40+ seconds to save or publish. The only thing I can gather is the "PostSave" in the network tab stays on pending for at least 40 seconds before the document saves. The document in question only has two fields, a title and a richtext editor with barely any content.
I've ran the code locally and saving the same documents takes only a second or two.
I've seen others with the same problem but haven't found a working solution yet. Any ideas?
Should mention we've imported some content so the site has around 24000 documents published,
Does your instance have event handlers doing stuff on save/publish events?
There could be a any number of checks, verification routines, email notifications being sent out..
What about something like (the super awesome)uSync that generates/updates files on disk on these events? It's possible to configure your site in combination with these types of functionality in such a way so as to trigger File Change Notification and cause an app restart (easy to tell from the logs).
These are just guesses from a million miles away, of course, but the first two spots i'd check in the processes of elimination. -good luck!
Just wanted to do a (quite late) update, but we did eventually discover that we had a document type that iterated through ALL nodes on publishing. I'll mark your answer as correct since
Does your instance have event handlers doing stuff on save/publish events?
Saving certain documents in the backoffice is very slow
We're running a 8.17.2 fairly fresh umbraco site on a 2019 windows server with a MSSQL db. Noticed lately that certain documents, seemingly random ones, will take 40+ seconds to save or publish. The only thing I can gather is the "PostSave" in the network tab stays on pending for at least 40 seconds before the document saves. The document in question only has two fields, a title and a richtext editor with barely any content.
I've ran the code locally and saving the same documents takes only a second or two.
I've seen others with the same problem but haven't found a working solution yet. Any ideas?
Should mention we've imported some content so the site has around 24000 documents published,
Does your instance have event handlers doing stuff on save/publish events? There could be a any number of checks, verification routines, email notifications being sent out..
What about something like (the super awesome)uSync that generates/updates files on disk on these events? It's possible to configure your site in combination with these types of functionality in such a way so as to trigger File Change Notification and cause an app restart (easy to tell from the logs).
These are just guesses from a million miles away, of course, but the first two spots i'd check in the processes of elimination. -good luck!
Just wanted to do a (quite late) update, but we did eventually discover that we had a document type that iterated through ALL nodes on publishing. I'll mark your answer as correct since
Is what ended up being the cause.
is working on a reply...