It’s Monday morning when a company contacts me. Since last Friday, they have stopped receiving emails on all their email addresses. They ask me if I can investigate and solve it. The company has a subscription with Antagonist for hosting and domain name. They use the IMAP protocol.
I send an e-mail to one of the e-mail addresses and the company says that it has not received the e-mail. Strange because I don’t get an email back that the inbox is full or is seen as spam.
The next step is that I log in to mijn.antagonist.nl. In the overview I see that they are close to the limit of 50 GB. There is almost no space left. So that will be the reason emails stopped arriving. But the e-mails are parked somewhere because Antagonist’s mail servers do not return e-mails.
I can see in the graphs from last week that on Friday the space consumption went from 27 GB to 50 GB in less than 24 hours.
I log into Antagonist’s DirectAdmin environment so that I can see if:
- The website suddenly added a lot of files
- The database has grown spontaneously
- The mailboxes have grown very fast
The conclusion is that the problem is with one of the mailboxes that went from 2.7 GB to 27 GB. This corresponds to the graph from step 1. I can also see which mailbox it is in the overview of mailboxes.
The user of the mailbox indicates that he cannot explain why he started using 10x as much space in 24 hours. With his permission I therefore log in via the webmail environment. I don’t see a lot of folders. I can therefore easily check them 1 by 1 and add up to about 2.7 GB.
It is clear that there is a space problem that is specific to 1 mailbox. I therefore decide to log in with Filezilla on the server and examine the folder of the mailbox.
In the following folder:
/imap/<< domain name >>/<< user name >>/mdbox/storage
I find thousands of files that:
- Made within seconds of each other
- Have a sequential name of m.3000, m.3001, m.3002
- Have the same size in bytes
- Made on that Friday when the use of space has grown enormously
I randomly download a few files to compare against each other. These are identical. It is a specific email.
In consultation with the owner of the mailbox, I delete all versions of the same file from the server with the exception the first file. I consider the first file to be original.
After this I scroll further through the “storage” folder and find 1 more file that hundreds of times occurs with exactly the same 4 properties as shown above. I also delete all versions of this except the first one.
So the solution was to delete the thousands of duplicate files.
Within 1 hour the graph on mijn.antagonist.nl and DirectAdmin has dropped back to the number it was on Friday. The consumption is therefore slightly more than 50% of the total, which is fine.
Suddenly all e-mails sent after Friday also arrive in the mailboxes. This also applies to my test email.