IP address registered in Plesk is broken...

[VanyaF]

Hello,

The problem happens frequently, one of the IP addresses that is added to the server and has the SSL just stop working. In order to fix this issue I just went into "Server" section in the Plesk control panel --> "IP addresses management" and clicked on the "Repair" button, but now the problem happens too regularly and I've already got a lot of the emails from my customers
In the admin panel logs I just found the following information:


2012-07-31T22:14:16-03:00 ERR (3): Execution failed.
Command: ifmng
Arguments: Array
(
[0] => --add
[1] => eth0
[2] => 186.202.164.110
[3] => 255.255.252.0
)

Details: Empty error message from utility.

My OS and panel version:

cat /usr/local/psa/version
11.0.9 Debian 6.0 110120608.16

Also when I restarted the networking the system defined the IP address and I was able to view it in the output of the command "ifconfig" but IP address was still marked as broken in the Plesk
Please assist me in the resolution of the problem, any information would be very appreciated

 

[IgorG]

What sort of output

# /usr/local/psa/admin/bin/ifmng -l

?

 

[VanyaF]

Currently the information is correct since I've already repaired the broken IP address:

root@cpro12164:~# /usr/local/psa/admin/bin/ifmng -l
186.202.164.108 255.255.252.0 eth0 1
186.202.164.110 255.255.252.0 eth0 0
186.202.116.41 255.255.255.0 eth0 0

I would like to know your ideas regarding the reason of the problem and where to "dig" in order to investigate and troubleshoot it

 

[IgorG]

Check records in psa database:

mysql> select * from IP_Addresses;

 

[VanyaF]

Currently I'm getting the following results:

mysql> select * from IP_Addresses;
+----+---------------+-----------------+---------------+-------+--------------------+-------------------+-------+-------+--------+
| id | serviceNodeId | ip_address | mask | iface | ssl_certificate_id | default_domain_id | ftps | main | status |
+----+---------------+-----------------+---------------+-------+--------------------+-------------------+-------+-------+--------+
| 1 | 1 | 186.202.164.108 | 255.255.252.0 | eth0 | 2 | 0 | false | true | 0 |
| 2 | 1 | 186.202.164.110 | 255.255.252.0 | eth0 | 5 | 4 | false | false | 0 |
| 3 | 1 | 186.202.116.41 | 255.255.255.0 | eth0 | 6 | 13 | false | false | 0 |
+----+---------------+-----------------+---------------+-------+--------------------+-------------------+-------+-------+--------+

To which details should I point my attention first of all ?
Thank you for your assistance and cooperation

 

[abdi]

Vanya, everything looks fine probably because of the repair you already made. However may I suggest you run that SQL when the problem comes back again?

 

[VanyaF]

Certainly, I'll update this post once the problem comes back

Thank you

 

[abdi]

Sure no problem! we will wait to hear from you.

 

[johnrdorazio]

I have also been having this problem quite frequently, I am having to manually "repair" from the "Plesk -> Server -> IP Addresses" page an IPv6 address on my server.
It happened again today, and I saw that the address is defined in "/etc/network/interfaces", but does not show up in "/usr/local/psa/admin/bin/ifmng -l". It does show up in the database "psa -> IP_Addresses", but while all other addresses have status 0, the broken IPv6 address had status 1 (I have no idea what that means).

Every time this address is "corrupted" it stops my nginx with an error of not being able to bind to this address. I can only restart nginx after repairing the address from the Plesk interface.

It would be great to fix this permanently, but even if I could just automate the process of repairing and restarting nginx from a bash script would be better than having to do it manually seeing that all my websites on my server go down for hours on end at times becaues of this.

I would like to know what exactly Plesk does when it "repairs" an IP Address? How can I do the same thing from the command line, instead of from the web interface?

 

[IgorG]

@johnrdorazio

Try to apply solution from http://kb.odin.com/en/114245

 

[johnrdorazio]

@IgorG
In that KB page it says to click the "repair" link in the Plesk interface IP management page, I already do this and it works for me (for some reason it happens often).

1. ...
2. You can use the following steps to repair the IP address:
   
   a. Access Tools & Settings > IP Addresses
   
   b. If there is an IP address marked with an exclamation point and a broken state, click the repair link.
   
   c. If there are no broken IP addresses, click re-read. Now, these broken addresses should be revealed and repaired as per p.2

Instead I would like to know what the "repair" link is doing behind the scenes. How can I do the same thing from command line? This way I can detect from a bash script in cron when my nginx status is down, and automatically issue the same commands from the bash script to "repair" the IP address...