Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a problem. My developers are using one of my local servers that is located in our office. The server is running PostgreSQL and when they try to make a query to it, it just lags and then gives out a timeout error. Then I started to ping it and trying to make querys my self. When ping is running and the server is responding to the pings PostgreSQL works just fine. But its not only the postgresql. When I am at home everything is working just fine. When I am at office it starts to play around and drop the connections. Some people even can not ping the server it gives them timeout. SSH'ing in to it from office is a pain. Sometimes it works and sometimes it does not because the server does not respond, but everything in the server is working fine.

Things I have checked for solutions:

  1. dmesg on server ( no hardware errors )
  2. Kernel log ( no errors )
  3. Factory Reseted all the routers and setup them again ( maybe route tables were off, but did not work )
  4. Change the Ethernet cable on the server ( no luck )
  5. Rebooted ( no luck )
  6. Nothing is eating away the resource of the server ( over 20GB ram left to use, CPU 0.1% )
  7. IP tables on the server ( everything is in order, even tried to reset them, no luck )

Can anyone help me with this? I am running out of ideas what could be the problem.

Thank you.

share|improve this question
1  
Are you sure you don't have a second piece of equipment configured with the same IP address? –  MadHatter Apr 25 '13 at 11:24
    
I think it could be also a firewall related issue. What FW are you using on that server and also could you put us your INPUT policy table? –  Str82DHeaD Apr 25 '13 at 13:51

1 Answer 1

up vote 1 down vote accepted

You might have a duplicate MAC address in your network which matches the MAC address of your server's network interface. When this happens, every time the station with the duplicate MAC address is sending a frame, the switches re-learn it to their forwarding tables and send traffic intended for your server to the rogue station. Running ping makes sure your server is frequently sending frames, triggering the re-learning into its favor.

As the issue seems to be connected to "you being in the office", it seems likely that a device which is only switched on at business hours is the culprit. You might try changing the MAC address of your server to a known-unique one or creating a static entry in your switches' FDBs (only possible with managed switches) as a remediation measure.

share|improve this answer
    
I like this idea, along with @MadHatter. Can you try to isolate the server and a single client to test connectivity –  David Houde Apr 25 '13 at 11:29
    
I will try both of them :) Thanks guys. –  mYzk Apr 25 '13 at 11:43

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.