Why is the mail queue processing emails for siteworx account that has reached their disk quota limits


We’re sitting with an issue, we have a few clients that have reached their hosting quota limits but they still receive emails from their clients. These emails still end up in the mail queue waiting to be processed. This, however, causes an issue and makes the mail queue “stuck”. It prevents the processing of incoming emails and fills the queue up.

We need to manually delete emails sent to these domains that have reached their quota limits, for the queue to process itself again. Why are no bounce emails sent to sender addresses instead? Why still queue the email if their disk space is full?

Increasing the disk space is the solution, yes. However, clients tend to take a while to consider upgrading their packages. Is there no way we can configure emails to bounce to senders if a recipient mailbox is full and not have it remain in the queue hampering other recipient addresses each time?


If the siteworx account disk quota is full then it cannot accept any more data

However depending upon the website parts of it may still work but the website may show issues

The email sending of website I believe is not counted in the disk quota of the siteworx account

Many thanks


Thank you for the feedback. But this is for receiving emails, not sending. The domains with full disk space keep getting emails, and they continue to remain in the queue and the server queue constantly tries to keep delivering it but cannot because there’s not enough space for said domains.

This process builds up and causes the queue to eventually get stuck, preventing other users, who have enough space not to get their emails.

How do we stop or configure the server for those who send emails to domains that are full to get bounce emails, instead of those emails ending up in the queue being constantly processed over and over?


The messages wait in the queue so that, when quota issue is resolved, those users get their messages. This is also, in my experience, often the only way that it is ever discovered a user IS over quota.

It is not possible to change this behavior–this isn’t anything we actually did on purpose. As far as I am aware, it is just how qmail handles storage usage.

You may want to set up the SiteWorx Storage Limits problem detector in NodeWorx under Server > System Health. That will send alerts if an account is at, or reaching, their quota, and you can then address the issue before it occurs: How to: Manage System Health Alerts — InterWorx documentation

Friendly Neighborhood InterWorx Support Manager

The process makes sense, in a way, yes. But would it not be better to send a bounce email to the senders saying user xxxx has reached the disk space limit try later and drop it from the queue?

As mentioned, this keeps piling up leading to the queue getting stuck and affecting other domains on the server that do not have disk space issues. At least this is happening with our instance recently.


Yes, I understand the behavior you would like, instead. There is not a way to change the behavior at this time, because this is not a decision that we made. We do not control this behavior. I simply don’t have a setting you can change on your server to do what you are looking for. Believe me, I wish we had a setting for that, because it is a common, and annoying, issue.

This is because qmail is simply not smart enough to really know “account abc is over quota, so I will just not deliver messages to account abc, but I’ll deliver to everyone else”. It just sees “an account is over quota. I don’t know what account it is, so I won’t deliver any messages right now until that is fixed, just to be safe”.

There is an existing ticket to see if there is a way to better work around this behavior, which I’m going to bump, but I do not have any guarantees, since, again, it’s just kind of how qmail does things. It has been a pain in my side for many, many years, so I 100% get your frustration.

We do have plans to hopefully be able to focus on switching to exim this year, which will, by nature, change how messages for over quota accounts are processed. That is just a goal, so I don’t have any kind of ETA, but that is in the pipeline.

For now, however, again, I recommend enabling the storage alert problem detectors, so that you are able to keep an eye on when an account may be reaching quota, so mitigate the issue before it occurs. That is really all that can be done at this time.


Thank you for the detailed feedback. Well noted and understood.