Users discuss rebaking posts in Discourse after installing plugins or restoring backups. bekircem installed the spoiler alert plugin and found old posts didn’t render spoiler tags, requiring a full rebake via rake posts:rebake — a resource-intensive process for 300k posts. After high CPU load, gerhard suggested flushing Redis to stop jobs, but bekircem reported the process persisted; gerhard then advised killing the rake process directly. bekircem later asked how to rebake only posts with [spoiler] tags, and gerhard recommended rake posts:rebake_match["[spoiler]"] and linked to meta.discourse.org. Gulshan_Kumar faced broken images after restoring a backup, wondering if rake posts:rebake equals rebuilding — codinghorror confirmed rebaking helps if URLs changed, but gerhard clarified that missing thumbnails stem from disabled include thumbnails in backups, which triggers background rebaking. Gulshan_Kumar enabled the setting, re-backed up, and restored — yet some images remained broken even after rebuilds, with rebaking still slow on 8 vCPUs/52GB RAM. gerhard also suggested rake posts:rebake_uncooked_posts to expedite image recovery.
I’ve installed spoiler alert plugin. It works correctly. But old post doesn’t work spoiler tag.
If you change this, you must rebake all posts with:
rake posts:rebake.
How can do that? When i ./launcher enter app rake is not installed!
I’ve 300k post. Even if I did, the process may take a long time. Is there a way to do it in a short way?
Either one of these should work
./launcher enter app
rake posts:rebake
or
cd /var/discourse
./launcher enter app
bundle exec rake posts:rebake
I applied that commands and after close terminal but it seems working now on the terminal with htop command.
bin/rake posts:rebake
But the server is high load because of that. There 300.000 post. I think it will take a long time. When this is done, spoiler tag for old post work correctly, right?
How can i cancel that rake process. Server is high cpu load because of that.
So, the rake task finished and put lots of rebake jobs into Sidekiq and now your server is on 
?
Hmm, I guess, you could stop all tasks by flushing redis. 
You may loose some other background jobs in the process!
cd /var/discourse
./launcher enter app
redis-cli flushall
After that, take a look at the following topic. It should allow you to rebake only posts that contain the spoiler tag.
I applied actions.
Output:
OK
Anyway, bin/rake posts:rebake process still looks terminal and high cpu load continues.
Well, then kill the rake process.
I killed process. Server load down 5+ to 1.6 but CPU running still %70+.
How should it be rebake posts matching pattern for containing tags?
Try flushing redis again.
Something like rake posts:rebake_match["\[spoiler]"] should work. You’ll have to try it out yourself…
Read the post about rebaking I linked above. The documentation is really good!
This process has high cpu usage. %50,%90
postgres 9.5/main: discourse discourse [local] SELECT
What does it do?
After restoring Discourse on a new host, I found some old topic images are not showing
But it appears in two case
If I do below command
cd /var/discourse
./launcher enter app
rake posts:rebake
Is it same as rebuilding? 
Yesterday, I tried it was taking so long time, so I stopped it in half-way.
Thanks & Regards,
Gulshan
If the URL to your site changed, yes, it’s a good idea to rebake.
Site URL is same as before. The original goal was to moving Discourse existing forum to another hosting.
The include thumbnails in backups site setting lets you configure if backups should include thumbnails or not. The setting defaults to off and leads to smaller backups. When you restore such a backup you will see the following message in the log:
Posts will be rebaked by a background job in sidekiq. You will see missing images until that has completed.
You can expedite the process by manually runningrake posts:rebake_uncooked_posts
Thanks for the head up. Actually, I was surprised my original disk usage was near 34GB out of 40GB. but I was getting backup file of only 2GB. I have enabled this settings now, and I will take backup again and then I will proceed to restore.
include thumbnails in backupsRebuild HTML is not fixing the problem, just make that area little larger than before. Ref: screenshot
Not sure, if this will fix. 
On 8 vCPUs, 52 GB memory, the rebake process is quite slow. Is there any alternative fix?
