I would expect that the "In the Trends" section would always list current hashtags that are frequently used.
Since the feature was introduced, the same hashtags are listed.
Actual I've updated my instance to version 3.1.3 but as described the feature never worked really.
I'm not the only one experiencing problems:
https://discourse.joinmastodon.org/t/trending-hashtags-broken/2523
Can I provide some more infos?
Still an issue for me as well. Thought it was fixed, but only temporarily.
Do you have a backlog of hashtags pending review? Because then it would be a different issue
All hashtags queued for review are done. No hashtags appear in the review page for months now and the admin dashboard doesn't update as well. It's just...stuck.
Iβm on 3.2.0 and having the same problem. I have no pending hashtags, and havenβt gotten the notification emails in weeks.
Sorry for the comment by DeepL translator.
My server is v3.2.0 and I have the same problem. When I was researching this issue, I noticed that the Scheduler::TrendingTagsScheduler, a recurring job in Sidekiq, was not running.
This is the output of journalctl -u mastodon-sidekiq -g TrendingTagsScheduler -r.
9ζ 01 14:23:46 kokkoro bundle[1032199]: 2020-09-01T05:23:46.941Z pid=1032199 tid=ovxv5sxdr INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)
9ζ 01 14:18:46 kokkoro bundle[1032199]: 2020-09-01T05:18:46.942Z pid=1032199 tid=ovxv5sxdr INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)
9ζ 01 14:13:46 kokkoro bundle[1032199]: 2020-09-01T05:13:46.962Z pid=1032199 tid=ovxv5sydn INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)
I see Sidekiq is pushing a task to the queue, but none of the threads are running this job.
The task is set as lock: :until_executed. I was wondering if there was some garbage left on Sidekiq from a job that had been executed in the past for this task.
On further investigation I noticed that there was a lot of stuff left in the Unique Digests section of the Sidekiq dashboard. I deleted all of these. Then the Scheduler::TrendingTagsScheduler starts to work.
9ζ 01 14:28:47 kokkoro bundle[1032199]: 2020-09-01T05:28:47.094Z pid=1032199 tid=ovxtela0j class=Scheduler::TrendingTagsScheduler jid=368c9cb006a71673880d488d elapsed=0.16 INFO: done
9ζ 01 14:28:46 kokkoro bundle[1032199]: 2020-09-01T05:28:46.934Z pid=1032199 tid=ovxtela0j class=Scheduler::TrendingTagsScheduler jid=368c9cb006a71673880d488d INFO: start
9ζ 01 14:28:46 kokkoro bundle[1032199]: 2020-09-01T05:28:46.931Z pid=1032199 tid=ovxv5sydn INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)
I'm not familiar with Mastodon or Ruby, so I don't know if this operation is correct. I hope my attempts will help to resolve this issue.
Sorry for the comment by DeepL translator.
My server is v3.2.0 and I have the same problem. When I was researching this issue, I noticed that the Scheduler::TrendingTagsScheduler, a recurring job in Sidekiq, was not running.
This is the output ofjournalctl -u mastodon-sidekiq -g TrendingTagsScheduler -r.9ζ 01 14:23:46 kokkoro bundle[1032199]: 2020-09-01T05:23:46.941Z pid=1032199 tid=ovxv5sxdr INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler) 9ζ 01 14:18:46 kokkoro bundle[1032199]: 2020-09-01T05:18:46.942Z pid=1032199 tid=ovxv5sxdr INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler) 9ζ 01 14:13:46 kokkoro bundle[1032199]: 2020-09-01T05:13:46.962Z pid=1032199 tid=ovxv5sydn INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)I see Sidekiq is pushing a task to the queue, but none of the threads are running this job.
The task is set aslock: :until_executed. I was wondering if there was some garbage left on Sidekiq from a job that had been executed in the past for this task.On further investigation I noticed that there was a lot of stuff left in the Unique Digests section of the Sidekiq dashboard. I deleted all of these. Then the Scheduler::TrendingTagsScheduler starts to work.
9ζ 01 14:28:47 kokkoro bundle[1032199]: 2020-09-01T05:28:47.094Z pid=1032199 tid=ovxtela0j class=Scheduler::TrendingTagsScheduler jid=368c9cb006a71673880d488d elapsed=0.16 INFO: done 9ζ 01 14:28:46 kokkoro bundle[1032199]: 2020-09-01T05:28:46.934Z pid=1032199 tid=ovxtela0j class=Scheduler::TrendingTagsScheduler jid=368c9cb006a71673880d488d INFO: start 9ζ 01 14:28:46 kokkoro bundle[1032199]: 2020-09-01T05:28:46.931Z pid=1032199 tid=ovxv5sydn INFO: queueing Scheduler::TrendingTagsScheduler (trending_tags_scheduler)I'm not familiar with Mastodon or Ruby, so I don't know if this operation is correct. I hope my attempts will help to resolve this issue.
Can confirm this workaround worked, so looks like we identified the issue. Now we need people writing a patch for it.

On further investigation I noticed that there was a lot of stuff left in the Unique Digests section of the Sidekiq dashboard. I deleted all of these. Then the Scheduler::TrendingTagsScheduler starts to work.
Hi! We've had this issue for months, as @koyuawsmbrtn just confirmed while I was typing :wink: we now have trends working again after months. Just wanted to add a thanks to you.
Most helpful comment
Sorry for the comment by DeepL translator.
My server is v3.2.0 and I have the same problem. When I was researching this issue, I noticed that the Scheduler::TrendingTagsScheduler, a recurring job in Sidekiq, was not running.
This is the output of
journalctl -u mastodon-sidekiq -g TrendingTagsScheduler -r.I see Sidekiq is pushing a task to the queue, but none of the threads are running this job.
The task is set as
lock: :until_executed. I was wondering if there was some garbage left on Sidekiq from a job that had been executed in the past for this task.On further investigation I noticed that there was a lot of stuff left in the Unique Digests section of the Sidekiq dashboard. I deleted all of these. Then the Scheduler::TrendingTagsScheduler starts to work.
I'm not familiar with Mastodon or Ruby, so I don't know if this operation is correct. I hope my attempts will help to resolve this issue.