[mpm_prefork:error] MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting

[mdreira]

Hi!

We have seen a problem that only happens in WP developer.

In a WP developer installation, when I export a template to our personal cloud (with a special plugin that we have) this error appears in app logfile:

[mpm_prefork: error] [pid 1] AH00161: server reached MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting

The symptom in app, is that it cannot finish the export of the template and it stays loading indefinitely.

You can see a screenshot of the log here (WP developer): https://share.getcloudapp.com/9ZuQrDQP

To test, I tried to reproduce the problem in a WP managed by you.

Here you can see a screenshot of the log (WP managed): https://share.getcloudapp.com/JruxzGOl

The error is not given and I can do export normaly.

It seems by the error message, that we needed to increase the MaxRequestWorkers settings. Well...

So, what we did was go to the file manager of the WP managed application, then we went to the "apache" folder and then we visualized the configuration file "mpm_prefork.conf". We saw this: https://share.getcloudapp.com/yAuyxEb6

The parameter "MaxRequestWorkers" was at 6. Which made us think that in the WP developer app it would be lower.

When we went to this same file within the WP dev app, we saw that the parameter "MaxRequestWorkers" was at 15 !!.

We do not understand what is happening. Why 6 is enough for WP managed and 15 is not enough for WP dev ?.

What we did was raise the "MaxRequestWorkers" parameter to 50, in WP dev.

We still have the same problem. Nothing changed. We don't know why cloudron sees a problem in MaxRequestWorkers, maybe it's something else.

On the other hand, I have to say we had no problem before updating to 6.3. We could do the export without any problem and we never need to see where the "MaxRequestWorkers" parameter was.

Of course we restarted the app every time we change a parameter.

If you can give us a hand it would be great. Thank you.

[murgero]

@mdreira said in [mpm_prefork:error] MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting:

MaxRequestWorkers

This may need to change at the app level during apache's install.

[mdreira]

@murgero Apache is installed by default in the WP developer app. I can only access that parameter from the file manager. But even though I change the parameter it seems that it does nothing. I will wait for @girish or someone in support to help me.

[murgero]

@mdreira yes I meant at the app level for cloudron, not wordpress. It's an apache config that can't be changed via other means except the apache config file itself. This needs to be done by the cloudron team I'd imagine or you can fork the app and submit the change for them to review.

[mdreira]

@murgero It has to be something different that is interfering, I don't think that "MaxRequestWorkers" parameter is really the problem.

I have no problem with WP managed, it is something that WP dev has.

[girish]

@mdreira I too think this is probably not related to MaxRequestWorkers setting. There is no specific reason why those settings are different in both the apps. IIRC, we adjusted it a bit for some performance numbers (because people were comparing the performance of Cloudron WP install and WP install via other panels). Is there a way for us to reproduce this or does this happen only in your app?

[robi]

If you want to increase performance, remove Apache from the equation.

[mdreira]

@girish If you want I can create an empty WP dev, only with the template export plugin. And give you access to it.

I can create a video for you so that you know how to reproduce the problem.

As you have access to my cloudron, you can look at the logs you want.

Tell me if that's okay with you.

We could continue with this issue in private and then post here how it has been fixed.

It has to be some WP dev configuration that you have changed later in Cloudron 6.3 because everything worked fine before.

[girish]

@mdreira yes, please, happy to check what is going wrong. Can you drop me a note on support@ with the information?

[mdreira]

@girish ok, done. Email sent to support@