When I tried to load my Roundcube today, found that it failed to load the inbox and instead had thrown the following error:
Server Error! (Ok)
Then, I tried searching the cpanel logs or the roundcube error log but found nothing. Then, I checked the Dovecot log located here:
/var/log/maillog
I found the following:
May 7 13:57:49 network2 dovecot: imap([email protected])<26343><cQG+qxX7MvdneMrv>: Error: Mailbox INBOX: mmap(size=351817308) failed with file /home/mellow/mail/mellowhost.com/shawon/dovecot.index.cache: Cannot allocate memory
This is happening because Dovecot caches the mail index in a file, once it tries to cache a lot of emails, it fails with a memory error. In those cases, you may remove the cache file and let Dovecot generate a new cache based on the latest mails. You may simply rm the file and see Roundcube is loading again:
Your customer has deleted the emails using Outlook/Imap client or may have archived them on their PC. But the space usage didn’t change in your Cpanel counterpart. What do to?
Solution
The issue arises because Dovecot Imap does soft delete when issuing delete command. They don’t necessarily get deleted, instead marked for expunge. You need to issue the expunge/purge command to get the storage cleared. In Cpanel, they usually do it on the daily upcp command they run. But if you are not running upcp daily, or would like to voluntarily do this, you can do this by running the following command:
/scripts/dovecot_maintenance
If you do not run upcp daily, then you can run the above command using a cron, daily to make sure IMAP deletions are expunged daily to reclaim storage.
The above command expunges all the accounts that require expunge. But if you want to expunge any single account, you may do so using the following command:
When someone tries to send you mail from gmail/outlook, they get a reply mail with the following errors:
550 Please turn on SMTP Authentication in your mail client. mail-lf1-f43.google.com [209.85.167.43]:42770 is not permitted to relay through this server without authentication.
Remote Server returned '550 5.7.368 Remote server returned authentication required to relay -> 550 Please turn on SMTP Authentication in your mail client. ;mail-oln040092255101.outbound.protection.outlook.com;(APC01-HK2-obe.outbound.protection.outlook.com) [40.92.255.101]:9693 is not;permitted to relay through this server without authentication.'
Error Solution
The error appears for several reasons. I will try to point out the most common ones I have faced for clients.
One, this domain is listed in /etc/remotedomains, while the domain is actually a local domain or shall use a local exchanger. As the domain is set not to use local mail exchange, hence the MTA is bouncing the mail back from receiving. To resolve this problem, you need to go to:
WHM >> DNS Zone Manager >> Search your domain that is having problem >> Manage
Now, click on the ‘Email Routing Configuration’ just beside the ‘Actions’ button as shown in the screenshot.
In the MX, make sure to set your local MX, and then select ‘Local Mail Exchanger’, then save.
If the domain was in remotedomains, and you did the above, and the issue is still not fixed, then you may want to run the following command to rebuild the remotedomains and localdomains file for Cpanel
/scripts/checkalldomainsmxs –yes
The second reason, the issue can appear, if you have defined two different MX and one of them, does not have the user, to whom the mail is trying to deliver. You should try to avoid such ambiguity. If you use multiple different MX, then you need to make sure, both of them accepts the same set of the user list.
When you try to open the terminal from Cpanel, it shows you an error in the red screen like the following:
The WebSocket handshake failed at 1:34:27 PM.
Solution
The error is appearing because Cpanel uses a socket to create a terminal window for you from the Cpanel. But this verifies the origin of your URL. If you are behind a proxy, then Cpanel won’t let the socket establish.
Most of the time, I have seen the user using Cloudflare behind the domain and using the same domain to access the Cpanel. Unfortunately, Cloudflare’s proxied IP won’t be able to create the websocket. Hence, you must use a domain or hostname or the server IP to access the Cpanel, to use Terminal.
Question: When I try to send mail using a PHP script in Cpanel, it automatically switches the SMTP hostname to localhost, even if I use a remote relay like sparkpost or mailgun. How to fix this?
Solution:
The reason behind the behavior is, by default Cpanel doesn’t let your customers use remote SMTP using PHP. This is controlled using a feature called ‘SMTP Restriction’. Go to WHM, log in using root, from Security Center select “SMTP Restrictions” and now, disable this. Your script shall now be able to send mail using remote SMTP like sparkpost/mailgun/pepipost.
You might see the following set of errors with Cpanel
When trying to remove the redirect from Cpanel
Apache detected an error in the Rewrite config. <pre>httpd_ls_bak: Syntax error in -C/-c directive: Include/IncludeOptional: Could not open directory /usr/local/apache/conf.modules.d: No such file or directory </pre> Please try again.
/scripts/rebuildhttpdconf generates an error like the following:
httpd: Syntax error in -C/-c directive: Include/IncludeOptional: Could not open directory /usr/local/apache/conf.modules.d: No such file or directory
Resolution
Previously, we reinstalled apache24 to solve the issue like the following:
yum reinstall ea-apache24
Although you may also reinstall the apache config runtime extension, that shall fix the issue as well.
If you are trying to install the Elasticsearch 7.* in CentOS and Cpanel, you are probably seeing your Elasticsearch is installed, but not starting with an error like the following:
java.lang.NoClassDefFoundError: Could not initialize class com.sun.jna.Native
at org.elasticsearch.systemd.Libsystemd.lambda$static$0(Libsystemd.java:23) ~[?:?]
at java.security.AccessController.doPrivileged(AccessController.java:318) ~[?:?]
at org.elasticsearch.systemd.Libsystemd.<clinit>(Libsystemd.java:22) ~[?:?]
at org.elasticsearch.systemd.SystemdPlugin.sd_notify(SystemdPlugin.java:115) ~[?:?]
at org.elasticsearch.systemd.SystemdPlugin.onNodeStarted(SystemdPlugin.java:126) ~[?:?]
at java.util.ArrayList.forEach(ArrayList.java:1511) ~[?:?]
at org.elasticsearch.node.Node.start(Node.java:998) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Bootstrap.start(Bootstrap.java:313) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Bootstrap.init(Bootstrap.java:408) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Elasticsearch.init(Elasticsearch.java:167) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Elasticsearch.execute(Elasticsearch.java:158) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.cli.EnvironmentAwareCommand.execute(EnvironmentAwareCommand.java:75) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.cli.Command.mainWithoutErrorHandling(Command.java:114) ~[elasticsearch-cli-7.15.1.jar:7.15.1]
at org.elasticsearch.cli.Command.main(Command.java:79) ~[elasticsearch-cli-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Elasticsearch.main(Elasticsearch.java:123) ~[elasticsearch-7.15.1.jar:7.15.1]
at org.elasticsearch.bootstrap.Elasticsearch.main(Elasticsearch.java:81) ~[elasticsearch-7.15.1.jar:7.15.1]
The error appears if your tmp directory is not writable and executable by Elasticsearch. You need to make two changes, to make Elasticsearch start for you.
First, edit the elasticsearch sysconfig file:
nano /etc/sysconfig/elasticsearch
and add the following line:
ES_TMPDIR=/tmp
Next, you need to mount /tmp with exec permission, which is normally mounted with noexec permission in Cpanel:
mount -o remount,exec /tmp
You should be done now, you may start your Elasticsearch:
service elasticsearch start
Please also remember, Elasticsearch needs at least 3GB of RAM to start. If you have lower RAM than that, you are probably hitting OOM killer of the system.
Cpanel incoming mails are failing, with an error in the exim_mainlog as following:
failed to open db file /var/spool/exim/db/ratelimit: permission denied
The error is appearing due to some permission issues with the exim db or the files are corrupted. These files recreate when the exim restart. Hence, we can do the following:
# delete the db files
rm -rf /var/spool/exim/db/*
# restart exim
service exim restart
# fix permission of exim spool
chown -Rf mailnull.mail /var/spool/exim
chmod 0750 /var/spool/exim
It is trying to look into a sub-sub version of openssl, which is 1h for PHP 7.4, but 1i is installed by default which is not an exact match with 1h, hence the error is appearing. To fix the problem, you can not just run yum update on openssl and fix this as you technically already have the latest version. To work with this, you need to first uninstall the ea-openssl rpm and install the require version. You may do this in the following way:
[root@stack4 yum.repos.d]# rpm -qa|grep ea-openssl
ea-openssl11-devel-1.1.1i-1.el7.cloudlinux.x86_64
ea-openssl-1.0.2u-1.el7.cloudlinux.x86_64
ea-openssl11-1.1.1i-1.el7.cloudlinux.x86_64
[root@stack4 yum.repos.d]# yum install ea-openssl11-1.1.1h-1.el7.cloudlinux.x86_64
Loaded plugins: fastestmirror, rhnplugin, universal-hooks
This system is receiving updates from CLN.
Loading mirror speeds from cached hostfile
* cpanel-plugins: 185.125.185.32
* cloudlinux-x86_64-server-7: xmlrpc.cln.cloudlinux.com
Package matching 1:ea-openssl11-1.1.1h-1.el7.cloudlinux.x86_64 already installed. Checking for update.
Nothing to do
Remember to use rpm uninstaller instead of yum remove, as yum also removes the dependencies, but with rpm you can skip the dependency uninstallation and patch it the above way.
