We have a site where we use lots of {% cache %} tags.
Now we almost constantly have a "cleaning template cache" task running, and they often get stuck.
Manual DB intervention and resetting the stuck task to status pending, will sometimes make it move again.
I've tried the various solution here, but nothing helps: DeleteStaleTemplateCaches job getting stuck
Also - there are no error entries in craft.log or phperrors.log when it gets stuck.
Looking at the craft_templatecaches table I find lots of duplicate looking entries:
select id,locale, cacheKey, expiryDate, path from craft_templatecaches;
...
| 685607 | nb | zDzBdD8bc1y4pf9Sx2HFyL3i24YypSp2sAq5 | 2015-09-08 08:23:21 | site:kompetansetilbud/ny-i-hr-rollen-i-trondheim-10013 |
| 685610 | nb | TMz4eh7FPiINSUixwkpXb0SlkHnZ3b5TSgXm | 2015-09-08 08:23:22 | site:kompetansetilbud/ny-i-hr-rollen-i-trondheim-10013 |
| 685611 | nb | HgLIHHgf56nM1ggOdTmaLlaylu5JeOeM1dNm | 2015-09-08 08:23:23 | site:kompetansetilbud/ny-i-hr-rollen-i-trondheim-10013 |
| 685612 | nb | SLHUKfUz2fF5Jmpy0kWPpvMxFBHH8PvlVJEt | 2015-09-08 08:23:25 | site:kompetansetilbud/ny-i-hr-rollen-i-trondheim-10013 |
| 685614 | nb | oEU6fe1tGc7MdI8WSQIXsf2wZZd9qmXG8P9H | 2015-09-08 08:23:26 | site:kompetansetilbud/ny-i-hr-rollen-i-trondheim-10013 |
...
Is this normal?
Anything else I can do to debug why the task gets stuck?
Update:
Using keyed and global caching reduced the size of the craft_templatecaches table from ~20k rows to ~300. However, this did not stop the task getting stuck.