Search results

I made mail repair without success. Problem is still exists.

Unfortunately problem comes back. This is what I found in maillog:

I set default_vsz_limit to 1G. On status I didn't see any interesting entries.

Every day I receive alerts about imap server problem. Dovecot stops work for 2-3 minutes. In logs I have something like this: I don't know what is the problem source. Maybe too many connections etc.? Could you tell me how to debug this issue?

No errors in this log file. Today task manager have lower load about 50%. Maybe something big was to do.

Plesk task manager takes 100% all the time. I wonder how can I find why it's do that? psaadm 3358139 111 0.2 3083652 177736 ? Ssl 10:44 4:55 /usr/libexec/plesk-task-manager --config /usr/local/psa/admin/conf/task-manager.yml --data /var/lib/plesk/task-manager In longtask table I...

This is great answer. :) Conclusion is better to run many smaller servers (or VM) than a few big.

I would like to ask how many customers can I host on one server without issues. Let's assume that I have strong hardware like Epyc, much RAM and nvme storage. In this case can I host 100, 500, 1000 or more typical subscriptions with mail service? I'm wondering how much time server will do...

I have several servers with Plesk. When I reach about 200 subscriptions and around 1 TB of data backup process takes a long time like 10-15 hours. I wonder how to optimize backups? All Plesk installations run on Proxmox KVM VM. Servers are different configs but quite good like AMD Epyc or Xeon...

Version 18.0.59 fixed this issue.

No. I was greped all logs by new domain name. I think serial is not a problem. It's looks like Slave DNS extension don't send rndc query. On slaves server I don't see any sign about new domain. If my rndc key was wrong I won't update zone manually. It's right?

When I add new domain to Plesk server, DNS zone do not transfer to slave DNS server via Slave DNS manager extension. After manual update in domain DNS settings transfer is completing correctly. Here is full log from adding i manual update.

I suppose problem is in nginx check in /usr/local/psa/admin/sbin/modules/firewall/rules script. Try restart nginx before firewall apply and confirm. Chance for correctly confirmation is much more but not 100%. May be very long timeouts counted in minutes should workaround this issue.

This solution doesn't work. I have error message:

I don't think so is only my bad luck. I have 9 servers with Plesk on CentOS 7, Alma 8 and 9 and all of them with new firewall have this issue. Sometimes firewall reload correctly but I don't know when. I can't reproduce problem with 100% sure.

When I try start manually command: /usr/local/psa/admin/sbin/modules//firewall/rules --activate --rules-file /usr/local/psa/tmp/ext-firewall-rules-KhZNmg I have error message: Activation token is absent . This message appears much time before timeout.

It's looks like fail2ban doesn't matter. Is any other log except panel.log where I can see what is doing firewall?

Yes, timeout do not change anything. I can wait 1 second or 90 second with the same effect. Maybe problem is in another iptables app like fail2ban? What happens when fail2ban add something to firewall before confirmation?

I wonder why I have message like this: when in process list is: This process works to timeout from panel.ini. Is it process for firewall activation?

Yes, I confirm. :)