We went Live with our first Umbraco 8 (with 2 languages) this week. But the message under "Published Cache Status" makes me a bit nervous... Every day it changes to "Database cache is NOT ok (rebuild?)." I rebuild it every day manually so far.
Is this a problem? Or is this expected behavior?
What can I do to check why this is happening? Are there any lists or something I can check against? Or tutorials?
What do the numbers actually mean? How to tell if that is good or bad?
("ContentStore contains 619 items and has 4 generations and 1 snapshot. MediaStore contains 578 items and has 7 generations and 1 snapshot.")
Is it possible to trigger Database Cache-rebuild / Memory Cache-rebuild myself? Eg by having a job that does this every night? (I guess that shouldn't be the recommended way but as a temporary workaround)
I have seen the message during development also, but thought it had to do with all strange codes/changes in Dev-env.
The only "strange" thing I can think of we are doing is that we have a scheduled job that import/create new nodes/pages every day.
(Using ContentService with CreateContent and the SaveAndPublish or GetById if it's an update)
We also have members that register. (we have ca 45,000 members, a couple of new ones every day are created)
And then of-course editors who create/update pages.
Any help or suggestions would be greatly appreciated!
Having the same issue here. Can anyone point us in the right direction to troubleshoot this?
My setup is similar in that it has Hangfire scheduled tasks, and members signing up (very few, though). It seems to mostly happen on Live after a deploy from Dev. It doesn't always seem to result in visible issues, but sometimes nodes go missing and the cache has to be reloaded or rebuilt.
Since I wrote the post we have upgraded Umbraco a couple of times to (now) 8.16.0. This together with doing sync (creating/updating) content less often has resulted in us not seeing this problem for a long time.
We still have some (few) logs that complain about "failed to get content with id xxxx", but we haven't noticed any problems really. Guessing these problems are intermittent.
We have other problems though, not related to this. (I believe I have created a post on that too, slow member login/creation)
The only "strange" thing I can think of we are doing is that we have a scheduled job that import/create new nodes/pages every day. + the fact that you have 45K members.
Possible things that I think that can be releated (please make a DB backup before trying any of this):
Have you made changes in a doctype that you make changes in a field when selecting/deselecting: Allow varying by culture AFTER you already have created data with this doctype. In this case defenitly check out point 2 and 3.
Having too much old versions, or incorrect old versions (doctypes changes between versions)
By default when doing a SaveAndPublish it creates a new version and keeps an old version of the content. This quickly adds up and can create issues like the Save operation seriously slowing down, getting error to even getting data corrupted.
I would do a quick check to go to a content item that is saved a lot got to the i tab in the upper right corner and click on Rollback to see the number of items/versions that are stored.
With the ContentService and then command DeleteVersions you can make sure version older then x days/weeks/months gets deleted. Please let me know if you need help on this then I'll lookup my code.
According to this page -> https://umbraco.com/blog/umbraco-91-release/
the history cleanup will also be added by default to Umbraco 8.18 (basicly doing what Unversion did in Umbraco 7).
Having to much stuff in the Trash can, old version in the trash can especially after making changes to a DocType can cause issues, delete from trash when no longer needed to be safe.
There is an option to Publish with descendends which you can do on a root note to do a fresh Publish on all nodes.
Check the error log to get some insight on what content piece may be causing the issues.
Please let me know if any of these suggestions help. Or if you find another cause for this issue please also share.
Wow, very good pointers, thanks! I wish I had known about this a year ago. :-) It will be a nice list to check against next time look into the site.
Yes, we probably have.
It does have a lot of data saved in history, this is really not needed. So great tip! Will make sure to delete these. If you do have code it would of-course help. :-) But we will upgrade to the latest v8, so maybe we can delete it that way.
Published Cache Status not OK
We went Live with our first Umbraco 8 (with 2 languages) this week. But the message under "Published Cache Status" makes me a bit nervous... Every day it changes to "Database cache is NOT ok (rebuild?)." I rebuild it every day manually so far.
("ContentStore contains 619 items and has 4 generations and 1 snapshot. MediaStore contains 578 items and has 7 generations and 1 snapshot.")
Is it possible to trigger Database Cache-rebuild / Memory Cache-rebuild myself? Eg by having a job that does this every night? (I guess that shouldn't be the recommended way but as a temporary workaround)
I have seen the message during development also, but thought it had to do with all strange codes/changes in Dev-env.
The only "strange" thing I can think of we are doing is that we have a scheduled job that import/create new nodes/pages every day.
(Using ContentService with CreateContent and the SaveAndPublish or GetById if it's an update)
We also have members that register. (we have ca 45,000 members, a couple of new ones every day are created)
And then of-course editors who create/update pages.
Any help or suggestions would be greatly appreciated!
Umbraco 8.5.3.
Having the same issue here. Can anyone point us in the right direction to troubleshoot this?
My setup is similar in that it has Hangfire scheduled tasks, and members signing up (very few, though). It seems to mostly happen on Live after a deploy from Dev. It doesn't always seem to result in visible issues, but sometimes nodes go missing and the cache has to be reloaded or rebuilt.
Same problem here, does someone have an answer
Hello, were you able to solve this problem?
Thanks.
Since I wrote the post we have upgraded Umbraco a couple of times to (now) 8.16.0. This together with doing sync (creating/updating) content less often has resulted in us not seeing this problem for a long time.
We still have some (few) logs that complain about "failed to get content with id xxxx", but we haven't noticed any problems really. Guessing these problems are intermittent.
We have other problems though, not related to this. (I believe I have created a post on that too, slow member login/creation)
I'm trigger by the line:
Possible things that I think that can be releated (please make a DB backup before trying any of this):
Have you made changes in a doctype that you make changes in a field when selecting/deselecting: Allow varying by culture AFTER you already have created data with this doctype. In this case defenitly check out point 2 and 3.
Having too much old versions, or incorrect old versions (doctypes changes between versions)
By default when doing a SaveAndPublish it creates a new version and keeps an old version of the content. This quickly adds up and can create issues like the Save operation seriously slowing down, getting error to even getting data corrupted.
I would do a quick check to go to a content item that is saved a lot got to the i tab in the upper right corner and click on Rollback to see the number of items/versions that are stored.
With the ContentService and then command DeleteVersions you can make sure version older then x days/weeks/months gets deleted. Please let me know if you need help on this then I'll lookup my code.
According to this page -> https://umbraco.com/blog/umbraco-91-release/ the history cleanup will also be added by default to Umbraco 8.18 (basicly doing what Unversion did in Umbraco 7).
Having to much stuff in the Trash can, old version in the trash can especially after making changes to a DocType can cause issues, delete from trash when no longer needed to be safe.
There is an option to Publish with descendends which you can do on a root note to do a fresh Publish on all nodes.
Check the error log to get some insight on what content piece may be causing the issues.
Please let me know if any of these suggestions help. Or if you find another cause for this issue please also share.
Wow, very good pointers, thanks! I wish I had known about this a year ago. :-) It will be a nice list to check against next time look into the site.
Yes, we probably have.
It does have a lot of data saved in history, this is really not needed. So great tip! Will make sure to delete these. If you do have code it would of-course help. :-) But we will upgrade to the latest v8, so maybe we can delete it that way.
No, not too much trash.
4-5. Will do!
Hi, I have solved the problem by manually clearing the cache and republishing the site, I did a migration from version 7 to 8 and had this problem, this resource helped me in the process https://www.proworks.com/blog/archive/how-to-upgrade-umbraco-version-7-to-version-8 and the official umbraco documentation https://our.umbraco.com/documentation/Fundamentals/Setup/Upgrading/migrating-to-v8 I combined some of the steps and now I have the site migrated to 8.
Thank you very much, greetings.
is working on a reply...