PHP OPcache module is not properly configured: OPcache interned strings buffer is nearly full

[MyNameIsRatchet]

Belongs to this closed issue.

I'm using nextloud on docker ("production"-tag) with NC 23.0.5.
I'm seeing this "PHP OPcache module is not properly configured: OPcache interned strings buffer is nearly full, please enter a value higher than 16"-warning in my settings-overview.

I know that this setting has been changed from "8" to "16" in the past in this PR, but my instance wants to have it higher than 16.

Anyone else struggling with this?
Any help appreciated.

[cloudwindy]

Same here +1

[merzi]

+1

[Jerome-Herbinet]

Hello,

I confirm this.

I don't know if the problem mentioned in this message is real and the performance of the instance is reduced or if it is a misinterpretation of the situation (and thant there is actually no problem).

I also confirm the following:

• this problem has appeared on my instance since the very first version 23 (still here with Nextcloud 24.0.2)
• changing the value of the "opcache.interned_strings_buffer" parameter (with 16, 32, 64 ...) only solves the problem temporarily, then the warning message always comes back. This method has been tried by quite a few people and the problem seems to keep coming back.

Assuming that the problem cannot be solved right away, it would be nice to know if the problem is real (and potentially degrading the performance of the instance) ... and if there is a workaround while waiting for a lasting solution to the problem.

@CarlSchwan , have you heard about this and/or do you know anyone who is familiar with this part of Nextcloud?

Best regards,
Jérôme

[schroeder-hesel]

Same problem here. Using Version 24.0.3 with Docker.

[J0WI]

I think this is fixed in nextcloud/server#32902
For more details and recommendations, please follow the discussion in nextcloud/server#31223

[LouisOuellet]

This issue is still very much present

[henningsieh]

This issue is still very much present

As a temp. solution I did this successfully: nextcloud/server#31223 (comment)

I did a sudo service apache2 restart and it went away but then comes back after a while