Hi all,
Has anyone else noticed this error and is there a fix for it? I tried deleting the widget and adding again. Sometimes it will work for a session then revert back to the error again. Its the same for all websites.
.
Hi @Tony-Baker
Thanks for creating this issue. I've just checked the SEO widget, and for me t works without issues.
Are you already using the latest version of Matomo?
Hi @sgiehl ,
Thanks for the quick response, I really appreciate it. I'm using the latest version of Matamo 4.11.0, self-hosted via vultr. Everything else is working great. The issue seems to be intermittent. I can record a quick video to show you the issue if that helps?
I guess a video might not help much. Are you able to see any errors in the browser console? If not maybe you can check the result of the request, that tries to fetch the ranks from backend.
No browser console issues, the only thing that I spotted in the Diagnostics > system check is this:
This issue has been mentioned on Matomo forums. There might be relevant details there:
https://forum.matomo.org/t/seo-ranking-widget-intermittent-error/47105/2
For information, on the demo website, I tried this widget, and I had to press on the [Rank] button in order to get some results...
@sgiehl @MatomoForumNotifications @heurteph-ei . Demo site has the same issue:
Strangely it will work if you search on other websites and then press the 'Rank' button, just not the one currently selected for analysis.
This might be a random failure maybe. Matomo is caching the results, maybe even in case of an error. That's something we need to investigate...
any news on this @sgiehl we have same issue
As a side note deleting the tmp folder seems to fix it - though not ideal!
Sorry @craigvc. We currently don't have the capacity to work on that one. As mentioned before, this seems to be some sort of caching bug. In some cases when fetching the results a (random) error might occur. That result is currently cached and displayed over and over again till the cache is invalid or cleared.
We need to change how the caching is currently handled. I guess the smartest solution would be to lower the cache validity to something like 30 seconds in case of an error.
If someone is keen on looking into that one, feel free to provide a Pull Request. We would be happy to review and assist with getting it merged.