Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 433
  • Last Modified:

Loss of connection

We have a network with 5 Dlink DES 3526 switches. These managed switches were confiigured by a previous company, I have taken over there maintenace for about a year now, as the previous company lost the contract to support the network.
The network has been working fine, until last week when we started losing packets on the network.
The switches are configured so that they isolate 4 floors in the building.
We have been using a Vigor 2800 router to access the internet and provide dhcp to the network, 4 cables are connected from the router to 4 of the switches.
There is an sbs2003 server for 2 floors, the dhcp has been disabled on the server.
We are in a situation where users across the entire building are losing internet access, the users on the sbs2003 server are losing access to email, so outlook says its losing connection to the server.
At times when I ping the server, I get a reply and then packets are lost, how do I isolate the switch which is causing the problem?
I have considered testing by getting an unmanged switch and plugging the cables from the managed switch one after the other until I find which switch has the problem, is this a good idea?
Switches are not my strong point so bear with me.
0
goodadvice
Asked:
goodadvice
  • 3
  • 2
1 Solution
 
kyleb84Commented:
Your first issue is the fact you using D-Link (Not your choice I know).

From the Vigor 2800 I would see if it can give you any Ethernet stats on the 4 cable connections, CRC error stats etc...

If not, I would run a long continuous ping to each switch from the router - just to test the links there.

Are the switches all connected back to the router? Or are they connected to each other as well?
0
 
goodadviceAuthor Commented:
"From the Vigor 2800 I would see if it can give you any Ethernet stats on the 4 cable connections, CRC error stats etc..."

1) I am not sure what you mean but if I change the dhcp to the sbs2003 server and then pinged the switches, I might get an idea of which switch is mulfunctioning?

"Are the switches all connected back to the router? Or are they connected to each other as well?"

2) The switches are connected to the router, via the combo connections, I am not sure whether they are connected to one another as well, I am not in the office, but how could I tell.

0
 
plug1Commented:
This sounds indicative of a faulty network card or device on your network sennding out crap, it can be as simple as a bad connection on 1 cable. Luckily its very easy to trouble shoot, when you lose connectivity start a continuos ping from a pc on the first switch to the server and keep an eye on the time outs, then disconnect the other 3 switches 1 at a time, when the bad area is out connectivity should be restored (unless the pc is onthe bad switch if thats the case repatch and start again).  Once you have worked out which switch is causing the outage then move on to a pc on that switch patch it up and wait for connectivity to go again, when it does, remove each device 1 by 1 until it comes back up, now you have isolatde either a bad port on the switch or a bad device.
0
Efficient way to get backups off site to Azure

This user guide provides instructions on how to deploy and configure both a StoneFly Scale Out NAS Enterprise Cloud Drive virtual machine and Veeam Cloud Connect in the Microsoft Azure Cloud.

 
goodadviceAuthor Commented:
Thank you guys for the information, I will be awarding points to plug1.
What I realised after reading through your answer is that, I had introduced a powerline network device because we wanted to connect a printer to a point in the building where there were no network points.
I disconnected it from the network and immediatley everything was back to normal, I can confidently say that as the network has been up the whole day something that has not happened for over 5 days. After the relief I have not even tried to reintroduce it and test the network, but no packets are now being lost, and I plan to have a good nights sleep.

Now I don't know whether I have to start another question to find out which cisco switches we need to introduce into the network and if they have any mechanism to identify a problem like the one we had, or which would be the best way to be able to monitor the network so we know when something like that happens. Since some one in the building could plug a device on the network and before we know whats happening we would be back there again.
0
 
plug1Commented:
Good stuff, the thing with these problems are once they start they overload everything so quickly its pretty hard to mdiagnose it from packets etc. Disecting the network (or rememebring what youve just added lol) is the quickest way for me to find out when the sh1t hits the proverbial.
0
 
plug1Commented:
By the way you need to accept an answer to dish out the points. Cheers. :)
0

Featured Post

Upgrade your Question Security!

Your question, your audience. Choose who sees your identity—and your question—with question security.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now