Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: [1] 2
1
PHP / CWP internal PHP 7.1: intl extension fails after recent updates - missing libicu
« on: June 25, 2026, 06:51:39 AM »
Hello,
I have two different CWP8 servers on AlmaLinux 8.
Since June 24/25, both servers started showing the same warning in cron.daily/cwp and cwp_bandwidth:
PHP Warning: PHP Startup: Unable to load dynamic library 'intl'
libicui18n.so.50: cannot open shared object file: No such file or directory
This is not related to the PHP version used by websites. The warning comes from CWP internal PHP:
/usr/local/cwp/php71/bin/php
Checks on both servers:
/usr/local/cwp/php71/bin/php -m
shows the warning.
ldd /usr/local/cwp/php71/lib/php/extensions/no-debug-non-zts-20170718/intl.so | grep "not found"
returns:
libicui18n.so.50 => not found
libicuuc.so.50 => not found
libicudata.so.50 => not found
libicuio.so.50 => not found
Installed ICU libraries are only version 60 and 69:
libicu-60.3-2.el8_1.x86_64
libicu-devel-60.3-2.el8_1.x86_64
libicu69-69.1-4.el7.remi.x86_64
There are no libicu .so.50 files in /usr/lib64.
The latest dnf updates around that date were mainly kernel packages, so I am not sure if this was caused by a CWP update or by an internal PHP/intl dependency issue.
Is this a known issue with CWP internal PHP 7.1?
What is the recommended fix?
I prefer not to create manual symlinks from ICU 50 to ICU 60/69 because that may break ABI compatibility.
Thanks.
I have two different CWP8 servers on AlmaLinux 8.
Since June 24/25, both servers started showing the same warning in cron.daily/cwp and cwp_bandwidth:
PHP Warning: PHP Startup: Unable to load dynamic library 'intl'
libicui18n.so.50: cannot open shared object file: No such file or directory
This is not related to the PHP version used by websites. The warning comes from CWP internal PHP:
/usr/local/cwp/php71/bin/php
Checks on both servers:
/usr/local/cwp/php71/bin/php -m
shows the warning.
ldd /usr/local/cwp/php71/lib/php/extensions/no-debug-non-zts-20170718/intl.so | grep "not found"
returns:
libicui18n.so.50 => not found
libicuuc.so.50 => not found
libicudata.so.50 => not found
libicuio.so.50 => not found
Installed ICU libraries are only version 60 and 69:
libicu-60.3-2.el8_1.x86_64
libicu-devel-60.3-2.el8_1.x86_64
libicu69-69.1-4.el7.remi.x86_64
There are no libicu .so.50 files in /usr/lib64.
The latest dnf updates around that date were mainly kernel packages, so I am not sure if this was caused by a CWP update or by an internal PHP/intl dependency issue.
Is this a known issue with CWP internal PHP 7.1?
What is the recommended fix?
I prefer not to create manual symlinks from ICU 50 to ICU 60/69 because that may break ABI compatibility.
Thanks.
2
MySQL / MariaDB Password Reset Script Doesn't Work
« on: October 26, 2025, 10:11:03 PM »
I know this script worked perfectly before.
I've recently done several installations of CWP 8 (for various tests), and after installing CWP 8 without restarting, if we run the password reset script, it works. However, if we restart CWP 8 after installing it, it no longer works, and from what I can see, we lose access to CWP administration (and therefore, the MariaDB service is inoperative).
I've tested it with both the stock and upgraded versions of MariaDB.
I understand that, over time and with updates, the script has become "outdated."
This is the code that appears when we run it:
Code: [Select]
/scripts/mysql_pwd_resetI've recently done several installations of CWP 8 (for various tests), and after installing CWP 8 without restarting, if we run the password reset script, it works. However, if we restart CWP 8 after installing it, it no longer works, and from what I can see, we lose access to CWP administration (and therefore, the MariaDB service is inoperative).
I've tested it with both the stock and upgraded versions of MariaDB.
I understand that, over time and with updates, the script has become "outdated."
This is the code that appears when we run it:
Code: [Select]
[root@server ~]# /scripts/mysql_pwd_reset
Enter the new root password (at least 8 chars).
Or leave it empty if you would like to generate it.
Or press CTRL+C to abort and do not touch it
Enter MySQL root password (NO special characters): testpass9876
Shutting down any mysql processes...
/scripts/mysql_pwd_reset: line 14: [: -eq: unary operator expected
Redirecting to /bin/systemctl stop mariadb.service
/scripts/mysql_pwd_reset: line 26: [: -eq: unary operator expected
Redirecting to /bin/systemctl start mariadb.service
Resetting password... hold on
--------------
UPDATE mysql.user SET Password=PASSWORD('testpass9876'),Authentication_string=PASSWORD('testpass9876') WHERE user='root'
--------------
ERROR 1348 (HY000) at line 1: Column 'Password' is not updatable
Cleaning up...
/scripts/mysql_pwd_reset: line 14: [: -eq: unary operator expected
Redirecting to /bin/systemctl stop mariadb.service
Password reset has been completed
New MySQL root password: testpass9876
Remember to store this password safely!
/scripts/mysql_pwd_reset: line 26: [: -eq: unary operator expected
Redirecting to /bin/systemctl start mariadb.service
[root@server ~]#
3
Installation / Problems installing CWP 8 with MariaDB
« on: October 24, 2025, 08:00:16 AM »
This week I installed CWP 8 from scratch, and I had to do it several times on the same server.
My mistake for not noting and saving the error.
The error appears once installed. It installs MariaDB 10.5, and when you run a yum update, it gives a result similar to this (because I searched for it and it appeared):
Copied from other post but equal than my problem:
yum update
Last metadata expiration date check: 0:05:32 ago on Sat 17 May 2025 14:34:10 UTC. Mistake:
Problem 1: package MariaDB-shared-10.5.29-1.el8.x86_64 from mariadb requires MariaDB-common, but none of the providers can be installed
- cannot install the best update candidate for package mariadb-connector-c-3.1.11-2.el8_3.x86_64
- package MariaDB-common-10.5.27-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.28-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.29-1.el8.x86_64 from mariadb is filtered out by modular filtering
Problem 2: package MariaDB-shared-10.5.29-1.el8.x86_64 from mariadb requires MariaDB-common, but none of the providers can be installed
- cannot install the best update candidate for package mariadb-connector-c-config-3.1.11-2.el8_3.noarch
- package MariaDB-common-10.5.27-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.28-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.29-1.el8.x86_64 from mariadb is filtered out by modular filtering
(try adding '--skip-broken' to skip uninstallable packages or '--nobest' to use not only the best candidate packages)
I read the link mentioned and it takes us Here:
https://www.alphagnu.com/topic/23-upgrade-mariadb-1011-in-cwp-centos-7-centos-8-stream-almalinux-78-rockylinux-78/#comment-25
But it doesn't fix it. In my case, it never fixed it; I always had the same problem.
Finally, I was able to fix it (I don't remember exactly how; I was stressed because the production server had crashed).
Now, cwp detects two pending updates. I log in and none appear. Yum update indicates that neither does it. But I see that the latest kernel, although apparently installed, isn't activated even in GRUB. I'm not worried; it will be updated with the next kernel released.
The problem is mariadb in a clean install. I think the installation script should be changed. It's not easy to fix the problem of having version 10.5 and then having yum update not work.
My mistake for not noting and saving the error.
The error appears once installed. It installs MariaDB 10.5, and when you run a yum update, it gives a result similar to this (because I searched for it and it appeared):
Copied from other post but equal than my problem:
yum update
Last metadata expiration date check: 0:05:32 ago on Sat 17 May 2025 14:34:10 UTC. Mistake:
Problem 1: package MariaDB-shared-10.5.29-1.el8.x86_64 from mariadb requires MariaDB-common, but none of the providers can be installed
- cannot install the best update candidate for package mariadb-connector-c-3.1.11-2.el8_3.x86_64
- package MariaDB-common-10.5.27-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.28-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.29-1.el8.x86_64 from mariadb is filtered out by modular filtering
Problem 2: package MariaDB-shared-10.5.29-1.el8.x86_64 from mariadb requires MariaDB-common, but none of the providers can be installed
- cannot install the best update candidate for package mariadb-connector-c-config-3.1.11-2.el8_3.noarch
- package MariaDB-common-10.5.27-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.28-1.el8.x86_64 from mariadb is filtered out by modular filtering
- package MariaDB-common-10.5.29-1.el8.x86_64 from mariadb is filtered out by modular filtering
(try adding '--skip-broken' to skip uninstallable packages or '--nobest' to use not only the best candidate packages)
I read the link mentioned and it takes us Here:
https://www.alphagnu.com/topic/23-upgrade-mariadb-1011-in-cwp-centos-7-centos-8-stream-almalinux-78-rockylinux-78/#comment-25
But it doesn't fix it. In my case, it never fixed it; I always had the same problem.
Finally, I was able to fix it (I don't remember exactly how; I was stressed because the production server had crashed).
Now, cwp detects two pending updates. I log in and none appear. Yum update indicates that neither does it. But I see that the latest kernel, although apparently installed, isn't activated even in GRUB. I'm not worried; it will be updated with the next kernel released.
The problem is mariadb in a clean install. I think the installation script should be changed. It's not easy to fix the problem of having version 10.5 and then having yum update not work.
4
Installation / Continue the problem that, can't migration account from cwp8 to cwp9?
« on: February 08, 2025, 01:01:54 AM »
Hi,
In the past I was try to migrate one account from cwp8 to cwp9 but I can't, error.
In this moment, its possible made this migration?
Thanks and regards.
In the past I was try to migrate one account from cwp8 to cwp9 but I can't, error.
In this moment, its possible made this migration?
Thanks and regards.
5
E-Mail / Strange problem with mac apple mail and new email accounts
« on: October 05, 2024, 07:10:53 PM »
I have this problem:
I create an account, and from any computer that is an iMac, it does not connect me to the account and even blocks me in the firewall.
If I connect to that account from the same PC, but with Thunderbird, there is no problem.
The log I get is this:
Oct 05 20:05:37 imap-login: Info: Disconnected: Inactivity (auth failed, 1 attempts in 180 secs): user=<usermail>, method=PLAIN, rip=xx.xxx.xxx.xxx, lip=xx.xxx.xxx.xx, TLS, session=<etrKl7xxxxxxPl02k>
And it blocks it.
I've gone around and around, but I can't figure out what's going on.
Any ideas?
Thanks
I create an account, and from any computer that is an iMac, it does not connect me to the account and even blocks me in the firewall.
If I connect to that account from the same PC, but with Thunderbird, there is no problem.
The log I get is this:
Oct 05 20:05:37 imap-login: Info: Disconnected: Inactivity (auth failed, 1 attempts in 180 secs): user=<usermail>, method=PLAIN, rip=xx.xxx.xxx.xxx, lip=xx.xxx.xxx.xx, TLS, session=<etrKl7xxxxxxPl02k>
And it blocks it.
I've gone around and around, but I can't figure out what's going on.
Any ideas?
Thanks
6
SSL / Moving server site, not SSL
« on: June 29, 2024, 03:13:51 PM »
Well this happens to me:
Original server (server1), destination server (srv1) .
Both Almalinux8.
I configure everything on a new server (srv1), and once I have everything configured, I start cwp transfer to cwp account.
Perfect, correct transfer.
I change DNS entries (I have them in cloudflare), all to the new server and I change the mail and srv1 entries but with transparency (cloufdlare cloud not activated in those entries).
On a new server, reload webserver, I am going to renew the certificate along with the email address and...
Error;The domain does not point.
And I can't...
I ping mail.mydomain.com and it tells me the new server/ip (I don't have the san entries with the orange cloud so that they don't generate "problems" in the SSL renewal).
I try and try, always the same error:
The domain does not point.
But if I generate the main ssl, the site is navigated, but I can't do any san...always error, the domain does not point...
I'm desperate because my old server is about to expire and I don't want to renew it.
And....I performed this same operation last month on two other servers, renewing the certificate and renewing it without any problem.
What difference is there from one to the other?
I have followed the same steps, and both sites, on cloudflare, are the same (as far as "free" entries for san I am referring).
I do not know what to do...
Thank you...
Original server (server1), destination server (srv1) .
Both Almalinux8.
I configure everything on a new server (srv1), and once I have everything configured, I start cwp transfer to cwp account.
Perfect, correct transfer.
I change DNS entries (I have them in cloudflare), all to the new server and I change the mail and srv1 entries but with transparency (cloufdlare cloud not activated in those entries).
On a new server, reload webserver, I am going to renew the certificate along with the email address and...
Error;The domain does not point.
And I can't...
I ping mail.mydomain.com and it tells me the new server/ip (I don't have the san entries with the orange cloud so that they don't generate "problems" in the SSL renewal).
I try and try, always the same error:
The domain does not point.
But if I generate the main ssl, the site is navigated, but I can't do any san...always error, the domain does not point...
I'm desperate because my old server is about to expire and I don't want to renew it.
And....I performed this same operation last month on two other servers, renewing the certificate and renewing it without any problem.
What difference is there from one to the other?
I have followed the same steps, and both sites, on cloudflare, are the same (as far as "free" entries for san I am referring).
I do not know what to do...
Thank you...
7
Migration from other control panels / I cannot migrate from Alma8 to Alma9, but I can migrate to Alma8
« on: June 26, 2024, 11:49:13 AM »
This happens to me, I have a server, Almalinux8, I want to migrate to another, I install Almalinux and I can't, it gives me an error (which I will indicate below), but on the other hand, if I install Almalinux8 on the new server, I don't have any problem in the migration, from what I deduce it is a matter of Almalinux9, since the only thing that varies with the new server is the operating system it installs.
This is the error:
Thanks
This is the error:
Code: [Select]
[root@srv1 ~]# tail -50 /var/log/cwp/account_transfer.log
2024-06-26 11:41:28 export SSHPASS=************; /usr/bin/sshpass -e /usr/bin/ssh -p4444 -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -o LogLevel=quiet root@MYIPSERVER
2024-06-26 11:41:28 test -d /usr/local/cwp/ && echo 'true' || echo 'false'
2024-06-26 11:41:29 true
2024-06-26 11:41:29 /usr/bin/ssh-copy-id: ERROR: Too many arguments. Expecting a target hostname, got:
Usage: /usr/bin/ssh-copy-id [-h|-?|-f|-n|-s] [-i [identity_file]] [-p port] [-F alternative ssh_config file] [[-o <ssh -o options>] ...] [user@]hostname
-f: force mode -- copy keys without trying to check if they are already installed
-n: dry run -- no keys are actually copied
-s: use sftp -- use sftp instead of executing remote-commands. Can be useful if the remote only allows sftp
-h|-?: print this help
2024-06-26 11:41:29 export SSHPASS=**********;/usr/bin/sshpass -e ssh-copy-id "-p 4444" root@MYIPSERVER -o StrictHostKeyChecking=no 2>&1
2024-06-26 11:41:29 /usr/bin/ssh -p 4444 -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -o LogLevel=quiet root@MYIPSERVER test -d /usr/local/cwp/ && echo 'true' || echo 'false' 2>&1
2024-06-26 11:41:29 false
[root@srv1 ~]#Thanks
8
Installation / ip route replace $(ip route list dev eth0...
« on: June 13, 2024, 05:00:30 PM »
I put it in this section, but I have no idea where to put it in the forum...
I recently installed cwp pro on a new VPS for me, Almalinux 8.
I migrated from my old VPS, everything is fine, but...
Every time I restart or after a few days I receive an email from the server:
Subject: Cron <root@server2> ip route replace $(ip route list dev eth0 scope link | head -n1 | awk '{ print $1 }') via $(ip route list dev eth0 | awk '/default/{ print $3 }') dev eth0
Content:
Error: any valid address is expected rather than "dev".
The server works well, I have not detected any problems but... it has me surprised, and with the "alarm set".
I've searched the internet but I haven't been able to see what's going on...
Does anything occur to you?
Thank you and greetings.
I recently installed cwp pro on a new VPS for me, Almalinux 8.
I migrated from my old VPS, everything is fine, but...
Every time I restart or after a few days I receive an email from the server:
Subject: Cron <root@server2> ip route replace $(ip route list dev eth0 scope link | head -n1 | awk '{ print $1 }') via $(ip route list dev eth0 | awk '/default/{ print $3 }') dev eth0
Content:
Error: any valid address is expected rather than "dev".
The server works well, I have not detected any problems but... it has me surprised, and with the "alarm set".
I've searched the internet but I haven't been able to see what's going on...
Does anything occur to you?
Thank you and greetings.
9
PHP / PHP Switcher version in Almalinux 9
« on: May 06, 2024, 08:14:09 PM »
I currently have two vps running on Almalinux 8.9, cwp pro.
As 8.9 reaches EOL soon, I have taken a VPS to do the migrations (VPSold to VPSnew, reconfigure VPS with Almalinux 9.x) and then, VPS new to VPS old.
Obviously, the new VPS that I use for this, I will do it from one account to one account, so I don't put CWP pro on it and I don't have access to PHP FPM (I don't need it).
My surprise was that PHP Switcher does not take me any version lower than php 7.4.33, I need php 7.1.33 but I can't.
Is there any limitation or am I doing something wrong?
As 8.9 reaches EOL soon, I have taken a VPS to do the migrations (VPSold to VPSnew, reconfigure VPS with Almalinux 9.x) and then, VPS new to VPS old.
Obviously, the new VPS that I use for this, I will do it from one account to one account, so I don't put CWP pro on it and I don't have access to PHP FPM (I don't need it).
My surprise was that PHP Switcher does not take me any version lower than php 7.4.33, I need php 7.1.33 but I can't.
Is there any limitation or am I doing something wrong?
10
PHP / 503 Service unavailable
« on: March 26, 2024, 06:47:35 PM »
I am encountering this problem and I don't know how to solve it.
I have a PHP script, which in the past worked for me without any problem on the same database, same environment...etc...
But it was on another host and another panel...
I launch it, it works, and from time to time, I get this error message, 503 service unavailable.
If I change a number in the url (what it does is that it imports data from one site to another) it continues without a problem.
After a lot of searching, the only error in the log I see is:
"child 4484 exited on signal 11 (sigsegv) after xxxxx seconds"
I have increased the times (set to -1) resources in php, in php-fpm I have set:
"pm = ondemand
pm.max_children = 100
pm.max_requests = 10000
pm.process_idle_timeout = 900s"
I did get that the max_children was low, it was advisable to raise it, but it was not the error.
I deleted the error log, I repeat the process, and the only error that appears is the previous one, signal 11...
I've been like this for 1 month or more, and I'm desperate...
Any ideas?
I have a PHP script, which in the past worked for me without any problem on the same database, same environment...etc...
But it was on another host and another panel...
I launch it, it works, and from time to time, I get this error message, 503 service unavailable.
If I change a number in the url (what it does is that it imports data from one site to another) it continues without a problem.
After a lot of searching, the only error in the log I see is:
"child 4484 exited on signal 11 (sigsegv) after xxxxx seconds"
I have increased the times (set to -1) resources in php, in php-fpm I have set:
"pm = ondemand
pm.max_children = 100
pm.max_requests = 10000
pm.process_idle_timeout = 900s"
I did get that the max_children was low, it was advisable to raise it, but it was not the error.
I deleted the error log, I repeat the process, and the only error that appears is the previous one, signal 11...
I've been like this for 1 month or more, and I'm desperate...
Any ideas?
11
PHP / Internal server Error or The server is temporarily unable
« on: March 06, 2024, 12:01:27 AM »
This happens to me.
I run a script in PHP, what it performs is a conversion of tables in the database, I used this script in the past, and it didn't give me any problems (in the same database) but... I used another panel.
Well, if I configure it to use php-cgi, at one point I get "Internal Server Error", if I configure it to use php-fpm, at the same point I get "The server is temporarily unable..."
And I've been dealing with this for many hours, I've tried everything I can think of and found, but I can't solve it.
Any idea? I'm starting to get a little desperate, to say the least...
PS: No errors in the apache or mysql log
I run a script in PHP, what it performs is a conversion of tables in the database, I used this script in the past, and it didn't give me any problems (in the same database) but... I used another panel.
Well, if I configure it to use php-cgi, at one point I get "Internal Server Error", if I configure it to use php-fpm, at the same point I get "The server is temporarily unable..."
And I've been dealing with this for many hours, I've tried everything I can think of and found, but I can't solve it.
Any idea? I'm starting to get a little desperate, to say the least...
PS: No errors in the apache or mysql log
12
SSL / AutoSSL and Cloudflare
« on: February 26, 2024, 02:51:39 AM »
Curious what happens to me, I have two servers, they are two different vps, both, the dns in the registrar point to Cloudflare and from there, to the sites.
Well, one, no problem, I have the A record of www and the site name pointing to VPS1 (proxyfied by Cloudflare, orange icon) and without problem, autossl renews me without any problem.
The second VPS, VPS2 and another site, there is no way, if I don't deactivate Cloudflare it won't generate the AutoSSL certificate... and I don't know what to do anymore, I'm a little desperate.
Thanks and regards.
Well, one, no problem, I have the A record of www and the site name pointing to VPS1 (proxyfied by Cloudflare, orange icon) and without problem, autossl renews me without any problem.
The second VPS, VPS2 and another site, there is no way, if I don't deactivate Cloudflare it won't generate the AutoSSL certificate... and I don't know what to do anymore, I'm a little desperate.
Thanks and regards.
13
CSF Firewall / The module csfofficial does not exist
« on: January 17, 2023, 11:43:32 AM »
Hello,
I have 2 servers, are same, same o.s. same version of cwp, same version of config server...al are same.
In one, I have in my menu: ConfigServerScripts - ConfigServerFirewall.
In the other server, I don't have this, and if I try to access by url per module, appears:
The module csfofficial does not exist
How can solved this?
Thanks and regards.
I have 2 servers, are same, same o.s. same version of cwp, same version of config server...al are same.
In one, I have in my menu: ConfigServerScripts - ConfigServerFirewall.
In the other server, I don't have this, and if I try to access by url per module, appears:
The module csfofficial does not exist
How can solved this?
Thanks and regards.
14
E-Mail / Not connect to mail from i-mac devices
« on: January 12, 2023, 11:01:28 AM »
Well, this strange case happens to me, from any other client, such as thunderbird, everything is perfect but... from two new i-mac devices, it does not connect to mail, testing, I have seen that, if I deactivate Firewall then it does connect, if I activate firewall does not connect.
But the imap firewall ports are correctly open, 993, 465...etc...otherwise it wouldn't work from Thunderbird for example.
And I don't like having to have the firewall disabled.
Any ideas? Because for me, it's the strangest thing I've ever seen...
Regards.
But the imap firewall ports are correctly open, 993, 465...etc...otherwise it wouldn't work from Thunderbird for example.
And I don't like having to have the firewall disabled.
Any ideas? Because for me, it's the strangest thing I've ever seen...
Regards.
15
Installation / Backup cpanel to CWP?
« on: February 09, 2022, 02:29:31 PM »
Hello.
I transfer some sites from cpanel to CWP trough cPanel-->CWP Migration, in this case, CWP contact direct with server with cpanel.
But now, I have a old backup of a old site and I want re open this site, I have a backup of this old account of cpanel, but not have this format: cpmove-{USER}.tar.gz
The format are:
backup-3.9.2018_16-47-07_xxx.tar.gz
This format is valid for option cpanel account restore?
Thanks and regards.
I transfer some sites from cpanel to CWP trough cPanel-->CWP Migration, in this case, CWP contact direct with server with cpanel.
But now, I have a old backup of a old site and I want re open this site, I have a backup of this old account of cpanel, but not have this format: cpmove-{USER}.tar.gz
The format are:
backup-3.9.2018_16-47-07_xxx.tar.gz
This format is valid for option cpanel account restore?
Thanks and regards.
Pages: [1] 2
