Resolved Cannot ping outside plesk server

[Net @ Jour]

Hello there,

After changing for a brand new server with centos 7 and plesk obsidian with IONOS, I cannot ping (for example with my windows cmd) the ip address server or any of the domains hosted on the server. www or ftp or directly the ip address any response.

It's crazy because the website are working fine and I can also connect on ssh to the server.
If I ping with SSH acces It works.

I do verify the firewall seems to be ok.
Plesk interface is ok te be reached also.

Bind is active but when I check service named-chroot status -l I obtain :

[root@8ec4424 ~]# service named-chroot status -l
Redirecting to /bin/systemctl status -l named-chroot.service
● named-chroot.service - Berkeley Internet Name Domain (DNS)
Loaded: loaded (/usr/lib/systemd/system/named-chroot.service; enabled; vendor preset: disabled)
Drop-In: /usr/lib/systemd/system/named-chroot.service.d
└─respawn.conf
Active: active (running) since Mon 2019-11-18 07:12:45 UTC; 12min ago
Process: 11114 ExecStop=/bin/sh -c /usr/sbin/rndc stop > /dev/null 2>&1 || /bin/kill -TERM $MAINPID (code=exited, status=0/SUCCESS)
Process: 14795 ExecReload=/bin/sh -c /usr/sbin/rndc reload > /dev/null 2>&1 || /bin/kill -HUP $MAINPID (code=exited, status=0/SUCCESS)
Process: 11131 ExecStart=/usr/sbin/named -u named -t /var/named/chroot $OPTIONS (code=exited, status=0/SUCCESS)
Process: 11126 ExecStartPre=/usr/sbin/named-checkconf -t /var/named/chroot -z /etc/named.conf (code=exited, status=0/SUCCESS)
Main PID: 11132 (named)
Tasks: 5
Memory: 147.8M
CGroup: /system.slice/named-chroot.service
└─11132 /usr/sbin/named -u named -t /var/named/chroot -c /etc/named.conf -u named -n 2

Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: couldn't add command channel 127.0.0.1#953: file not found
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: configuring command channel from '/etc/rndc.key'
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: couldn't add command channel ::1#953: file not found
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: reloading configuration succeeded
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: reloading zones succeeded
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: all zones loaded
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: zone 189.165.82.in-addr.arpa/IN: loaded serial 1574061849
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: running
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: zone lathermaledegestion.com/IN: loaded serial 2019111803
Nov 18 07:24:10 8ec4424.online-server.cloud named[11132]: zone lathermaledegestion.com/IN: sending notifies (serial 2019111803)

I Also made a plesk repair dns and a plesk repair all - without any success I connot ping outside the server.

Must I verify another services ?
How to make it works correctly.

Sorry for my english

Thank you

 

[Bitpalast]

Do you have an extra, external firewall at your provider?

 

[Net @ Jour]

Hello

Thank you for helping
Yes There is a firewall with my provider but I think it is ok.
I have another server with 1and1.
Also with centos7 ans plesk obsidian and I can reach it from my internet connexion. Ping responding fine.
You can try to ping the ip 82.165.189.203
Thx

 

[Net @ Jour]

Hello,

Problem still there but I have more information to help to soluce it.
The server have really problems with resolving external hosts.

All websites hosted in the server have problems to resolve google.com when using contact forms with recaptcha.

Must I repare all bind configuration ?
Plesk repair dns do not find any problem.

Is it a centos7 configuration problem ? Plesk ?

Thanks for helping

 

[Net @ Jour]

Trying to find a solution

[root@8ec4424 ~]# cat /etc/resolv.conf
; generated by /usr/sbin/dhclient-script
nameserver 2001:8d8:fe:53:72ec::1
nameserver 2001:8d8:fe:53:72ec::2
[root@8ec4424 ~]#

Can it be the solution : my resolv.conf only shows IP V6 ?
When I ping (cmd on windows the IPV6 from the server :
2001:8d8:1800:1aa::1

The result is :

Envoi d’une requête 'Ping' 2001:8d8:1800:1aa::1 avec 32 octets de données :
PING : échec de la transmission. Défaillance générale.
PING : échec de la transmission. Défaillance générale.
PING : échec de la transmission. Défaillance générale.
PING : échec de la transmission. Défaillance générale.

So it'sin french but means :

failure of transmission. General failure.


I think this paramters are wring in mycentos7
How can I resolve it ?

 

[Ales]

...
The server have really problems with resolving external hosts.
...
Is it a centos7 configuration problem ? Plesk ?

This is a OS configuration issue, not Plesk's.

Can it be the solution : my resolv.conf only shows IP V6 ?
When I ping (cmd on windows the IPV6 from the server :
2001:8d8:1800:1aa::1

Testing from your location (office, home, ...) is likely to fail, as these are not public resolvers. The access is likely to be limited to the server's local network.

On the server itself, try 'ping6 2001:8d8:1800:1aa::1'. Besides, ping might be blocked, just try 'nslookup google.com' on the server and see if you can get a reply.

If there's an issue, refer to your providers documentation or contact their support to find out the correct resolvers. Or use public ones, or your own.

 

[Net @ Jour]

Hum ok thank you.

 

[learning_curve]

@Net @ Jour You mentioned 1and1 above (1and1 are now branded as IONOS). Are all of your affected servers with IONOS? If so, what servers are your running? Dedicated servers? Cloud Servers? VPS? A managed service? Do you, yourself, have full root access to them all? Are you running your DNS within Plesk or totally within IONOS? We're asking because we use IONOS cloud servers and have previously seen examples of the ping issues / temporary name server failures that you've posted, but all of which were solved within IONOS. FWIW All the issues were not releated to Plesk in any way, only to IONOS.

 

[Net @ Jour]

Hello
Thanks for tour message
It s a cloud server I manage with a full root access.
I manage another server (not a cloud one a dedicated one) also with Ionos without any problems.


Ionos gave me the ip address to connect the server and a second one ipv6 I add.

I manager the dns with ionos and on plesk Dns are slave.

So thank you so much. I ll contact Ionos.

 

[learning_curve]

...It s a cloud server I manage with a full root access. I manage another server (not a cloud one a dedicated one) also with Ionos without any problems. Ionos gave me the ip address to connect the server and a second one ipv6 I add. I manager the dns with ionos and on plesk Dns are slave. So thank you so much. I ll contact Ionos.

IONOS cloud servers work very well indeed, using both IPv4 & IPv6 addreses. However, we know from our own experience, that the only OS that we use (Ubuntu 18.04 LTS - we were an early adopter at the time with IONOS) was / is modified by IONOS themselves, so it's not exactly the same as the Ubuntu standard OS release. Can't advise you on any mods to CentOS 7 as we don't use it anywhere. The mods IONOS made were / are; using ifupdown instead of netplan. The mods were made by IONOS because "...netplan in systemd for Ubuntu is not compatible with our network infrastructure..." (sic) These mods did initially affect things and gave us very similar errors to those that you have described. However, in due course, IONOS fixed all of them all themselves, which was nice!

Seeing as you have an IONOS cloud server, have you tried simply re-booting the server after any big Plesk updates and re-checked? It's easy to take server snapshots on IONOS Cloud servers before any Plesk updates and you can then restore to that snapshot if needed. A re-boot, followed by ping and name server tests is very quick and easy to do on your cloud server and it was how we discovered it, as the temporary workaround back then. You may need to do two sequential re-boots (sometimes) if new OS mods by IONOS do turn out to be the temporary reason for the errors.

Plus just a quick question; Why use the Plesk DNS slave if you're fully setup correctly with IONOS? We don't use it because the IONOS DNS is quite efficient & quick n' easy to use now. The Plesk DNS slave doesn't really add any huge amounts of value, but it does add more potential for errors (Synching / Plesk IONOS incompatibility etc) It maybe worth doing a few of your own tests, on a two different domains. One with & one without the slave function and then looking at the results...

 

[Net @ Jour]

IONOS cloud servers work very well indeed, using both IPv4 & IPv6 addreses. However, we know from our own experience, that the only OS that we use (Ubuntu 18.04 LTS - we were an early adopter at the time with IONOS) was / is modified by IONOS themselves, so it's not exactly the same as the Ubuntu standard OS release. Can't advise you on any mods to CentOS 7 as we don't use it anywhere. The mods IONOS made were / are; using ifupdown instead of netplan. The mods were made by IONOS because "...netplan in systemd for Ubuntu is not compatible with our network infrastructure..." (sic) These mods did initially affect things and gave us very similar errors to those that you have described. However, in due course, IONOS fixed all of them all themselves, which was nice!

So HUGGGGGEEEEE Thanks your message help me to find the solution. You're right, it was a IONOS bug, the modification I've made in the Cloud Panel Firewall were not correctly stored. So it was a firewall problem but not the Plesk's one but the Ionos Cloud Panel Firewall.

Problem solved

Thank you all for help

 

[johann.liebert]

So HUGGGGGEEEEE Thanks your message help me to find the solution. You're right, it was a IONOS bug, the modification I've made in the Cloud Panel Firewall were not correctly stored. So it was a firewall problem but not the Plesk's one but the Ionos Cloud Panel Firewall.

Problem solved

Thank you all for help

hey i have the same problem. what did you do on the firewall setting in cloud pannel?