Receiving exim paniclog on srv01.myserver.com has non-zero size notice

Any idea what is causing this issue?

I thought it may have just been something that happened during the initial installation of the server, as I only found one entry in the paniclog, so I just took a backup of the file and deleted that line, so that I would stop getting the alerts.

However, I received an alert email again from exim paniclog as there is a new entry in the paniclog file. Any idea why this is happening or how to track down the root cause?

Here is a copy of the email I receive from the server:

Subject line: exim paniclog on srv01.myserver.com has non-zero size
Body: exim paniclog /var/log/exim4/paniclog on [srv01.myserver.com](http://srv01.myserver.com/) has non-zero size, mail system might be broken. Up to 10 lines are quoted below

Checking the /var/log/exim4/paniclog file, this is what I see:

021-04-14 19:35:03 1lWrqR-000B0S-Px 1lWrqR-000B0S-Px bad addresses found in headers
2021-04-20 22:45:06 1lZ5fe-000CGE-97 1lZ5fe-000CGE-97 bad addresses found in headers

This is not very helpful message, so I don’t really even know where it’s coming from or what it means. Any ideas?

Thanks in advance!

And in case this helps. I went back into the exim mainlog around the time those two entries occurred and found this:

2021-04-14 19:35:03 1lWrqR-000B0S-Px 1lWrqR-000B0S-Px bad addresses found in headers
2021-04-14 19:35:03 1lWrqR-000B0U-Qe <= <> R=1lWrqR-000B0S-Px U=Debian-exim P=local S=1281
2021-04-14 19:35:04 1lWrqR-000B0c-WE <= <> H=localhost (srv01.myserver.com) [127.0.0.1] P=esmtps X=TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM:256 CV=no K S=1533 [email protected]


2021-04-20 22:45:06 1lZ5fe-000CGE-97 1lZ5fe-000CGE-97 bad addresses found in headers
2021-04-20 22:45:06 1lZ5fe-000CGG-Cf <= <> R=1lZ5fe-000CGE-97 U=Debian-exim P=local S=1281
2021-04-20 22:45:06 1lZ5fe-000CGG-Cf == [email protected] <[email protected]> R=dnslookup defer (-1): host lookup did not complete

I would pay attention to this row

2021-04-20 22:45:06 1lZ5fe-000CGG-Cf == [email protected] <[email protected]> **R=dnslookup defer (-1): host lookup did not complete**

Just a guess, but check if gmail needs to whitelist your ip and/or change your admin email with a something you fully manage, not google, hotmail, yahoo and so on…

Actually, I think it’s these two rows, specially the <> text. I assume this means it’s empty? I don’t know enough about this.

2021-04-14 19:35:03 1lWrqR-000B0U-Qe <= <> R=1lWrqR-000B0S-Px U=Debian-exim P=local S=1281
2021-04-14 19:35:04 1lWrqR-000B0c-WE <= <> H=localhost (srv01.myserver.com) [127.0.0.1] P=esmtps X=TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM:256 CV=no K S=1533 [email protected]

I don’t think it has anything to do with the gmail address, as I see that line throughout the log as the server is constantly forwarding the various emails I have to that gmail account and they don’t result in the “021-04-14 19:35:03 1lWrqR-000B0S-Px 1lWrqR-000B0S-Px bad addresses found in headers” error. Also, no issue with email coming in and out through the server with gmail, so I don’t think it has anything to do with that. I just left that 3 line in the log line in case it had anything to do with it.

But I think it could be as you mention, something to do with the admin email. I was previously struggling with the admin email, as I was trying to add a forward in the ~/.forward file or the aliases file, but for some reason they were being ignored. Then I found on the VestaCP forum or somewhere else that the way hestia/Vesta is configured, it ignores this I guess. So the only solution I found, was I just had to configure and add an email user on HestiaCP for the hostname / domain I use for the server and give the root user an email account. Which I then set to auto forward all email to my gmail account and delete the local copy.

I do the same configuration for all the other domains and their associated email addresses on this server. And they all work fine. So I’m a bit confused. I wish I knew what this error message meant in the first place and then I could perhaps troubleshoot it.

well hey! Something interesting. I just happened to check my spam folder on gmail and found an email that was sent out by the server at the time of this error.

And yep, as expected, it’s that <>, it seems the email is trying to be sent to <> blank.

Here it is:

From: [email protected]

Subject line:
Mail failure - malformed recipient address

Body:

A message that you sent contained one or more recipient addresses that were
incorrectly constructed:

< >: missing or malformed local part

This address has been ignored. There were no other addresses in your
message, and so no attempt at delivery was possible.

------ This is a copy of your message, including all the headers. ------

To: < >
Subject: =?utf-8?B?Cg==?=
From: Hestia Control Panel <[[email protected]](mailto:[email protected])>
X-Priority: 3 (Normal)
CC:
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
X-Mailer: Php/libMailv1.3
Message-Id: <[[email protected]](mailto:[email protected])>
Date: Tue, 20 Apr 2021 22:45:06 -0700

Can't connect to MySQL localhost
ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/run/mysqld/mysqld.sock' (2)

Ugh. I think I have no found another issue to deal with. It’s like it never ends. So, well, one problem is not sure where [email protected]_server_hostname.com got configured, but that email probably won’t go anywhere.

@falzo - do you have any idea where this is set in HestiaCP?

Perhaps if I change wherever that is sent, then this email will be able to be sent properly and no more paniclog???!?!

Next thing I need to figure out, is why the heck the MySQL server not reachable. The server also hasn’t had any issue and as far as I know, MYSQL is running just fine and has not crashed. So not sure what this is. Maybe just a timeout at some point?

there is no address given to where the mail should be sent. so obviously that’s a ‘bad header’ and triggers the paniclog. that subject line also looks weird.

maybe with sending from your gmail while forwarding to it you create a loop at least in case it bounces or is not deliverable for whatever reason? I’d suggest checking the mailqueue for frozen messages.

the hestia notification mail is probably supposed to go to your admin account / email, did you supply something for it and if so is this a real one?

for mysql: yes most likely just a timeout thing, if it does not occur regularly rather ignore it, could have happened during a longer import or longquery or on a network blipp or whatever.

1 Like

Thanks Falzo! Yes, this has only happened twice during all the configurations, migrations, etc. So I think it’s probably just a one off. However, I think I need to fix this issue with the empty email.

This appears as though it’s an email sent by HestiaCP though. So I assume this is something the control panel is trying to send to the admin of the server. I have looked through the Control panel. Where is this admin email set/configured? I don’t remember setting anything at the time of installation and I can’t find where I can set it now, in the CP interface. But I assume once I set it, this issue will be resolved.

Can you advise where this value is set?

Thanks!

I went through the entire control panel interface and could not locate where I would set a contact email for notifications from the server.

Under the general settings page: cog wheel → Configure, I could not find anything there. But I did find an option where I can select a checkbox to receive notifications when HestiaCP is updated. So what email are these updates going to go to?

Email of the admin user should be used

Thanks for clarifying that @eris. So I confirmed that under the admin user, my gmail account is listed. It’s listed as both the catchall address, as well as on the root user address (as a forward all email address).

This same configuration works just fine on all the other domains on this site, so no issue with gmail receiving email or being used. And the issue is that in the error log, it appears as though, there is no email set, as the email is blank. So I assume whatever script that’s run by HestiaCP is throwing in the empty email address. But as I mention, the email is added in both locations.

Any ideas?

One other thing to share. When I setup this server last week, I was migrating from VestaCP. I had done a backup of the admin user from the vestaCP server, the same way I did with the other primary user account on that server (I had two accounts, admin, and a regular user (which is where I register all the domain names used on the server).

I used the v-restore-user and the admin backup file from the vestaCP server. Could this have caused an issue with the admin user and what email is stored? During the backup, is there any local storage into a file for the email address that I can check to see if it’s causing an issue.

The issue is definitely that the hestiacp script is using a blank email address, and I can’t seem to understand why. thanks!