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

  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 510
  • Last Modified:

Network Unmanaged Switches...!!

Hello Experts,

Recently we have been experiencing power cuts problems at our building. When power is off for longer than 10 minutes all UPS go off and hence all the servers, switches, routers etc.

Set up is like this: We have Server running Ms Windows server 2003 and other small applications like ms access database, active directory etc. We also use this server as a file server. The server is connected to total of 4 D-Link 1024 switches, the client computers then connects to these switches...(flat network). All the client computers access the server this way. These four d-link switches are connected up using a patch cable that run from one switch to another...on some of the switch ports we have connected a total of 8 Wireless Access Points (mix of D-Links and Zyxel). That is how our network is set up...

The problem: When power comes back on (after being off for like 15 minutes or so) and all devices turn on at the same time...the network seems not be working. We troubleshoot and realise one of the switch is not working...meaning all users connected to it are offline. However, if move all the users from the non-working switch to the working switch they all get connected online. At first, I thought the switch is broken...only to realise if I move the server and router (gateway to the internet) to the broken switch and add some users on it; these users all get connected. By doing this I managed to narrow it down as follows: All users connected to FIRST, SINGLE switch where Gateway and Server are connected can reach the Server and the Internet. But, if I connect a SECOND switch to the FIRST switch with a patch cable; all users on the SECOND switch are continue to get stay Offline. The situation repeat when I connects servers and gateway to the second switch then join it with the first switch.

We manage to identify a bad cable every time when this happen. So, I thought the problem is that particular cable...I isolate the cable, only to realise after couple of days when power goes out and come back, the same problem re-appear. Then I find another bad cable and isolate it. However, at this time when I plugged in the old bad cable is working very fine!! So, bad cable today turns out to be good cable after couple of days...and situation keep repeating. The source of the problems keep changing.

We would come in the morning to find everything is working perfectly, then around lunch time electricity goes...when it comes back network starts to misbehave!!. I am puzzled as to why this ONLY happens when Electricity go off...if it is a network loop shouldn't be causing problems all the time?

Please assist.

2 Solutions
The connection bridge to switch, was it with cross-over cable or straight? If it is not cross-over this may cause problems although it should work with straight through cables.

Please check if you are using cross over cables from switches to interconnect.

If the light is on, the next question would be whether the switch-to-PC connection connected at 100-full, 100-half duplex, 10 full, or 10-half duplex. A duplex mismatch will cause communication to fail...sometimes connectivity WILL work if it's mismatched, but performance will suffer.

Also, the switch has 'auto uplink' which means if you plug a straight-thru into it, it will automatically goto crossover mode. If you used a patch cable that was a crossover cable, it would work for the POE device, but not for your laptop.

The POE adapter also has to negotiate link speed and duplex, and it's getting it right since when you plug into it, it works.
Hi There

I would suggest you take a look at the following
Because of the power outage it would seem that your user Pc's starts up before the server,and then the server can't assign new IP's to them

I recommend that when the DHCP server is fully up and running let the Users Log off and On
so that IP's can be assigned to them

Who's Defending Your Organization from Threats?

Protecting against advanced threats requires an IT dream team – a well-oiled machine of people and solutions working together to defend your organization. Download our resource kit today to learn more about the tools you need to build you IT Dream Team!

TMAAAuthor Commented:
The bridge cable is and has always been a straight through cable. It has been working for a while only recently it when these power cuts start happening.

I will swap and see.

All the ports are on full 100 duplex.
TMAAAuthor Commented:
Thanks Vo1ty,

I have now disabled DHCP role on the Server and all DHCP is done on the Gateways (Cisco Router 1841).

Things are back to normal now...waiting until everyone leave office and test again.

Thank you.
Fred MarshallPrincipalCommented:
Depending on the equipment, arp cache updates can take a long time and create a very confusing situation.  Under the assumption that the "bad" switch is waiting to update, have you tried rebooting the "bad" switch before doing anything else.  Probably yes but I must ask, eh?-

Similarly, have you tried rebooting one of the computers that won't connect?  

If a cable is deemed "bad", it should be thrown in the trash once and for all.  Patch cables are "disposables".  Because of some of the things you've said, I would worry a lot about bad cables.  Poor workmanship on cable construction can cause a host of worries although probably not tied to power failures.  If ANY of the cables don't have the cable jacket firmly crimped into the connectors (so you can see the individual wires coming out the back of the connector), either replace those cables with manufactured cables or re-terminate them (put on a new plug) using truly good technique.  Bad cables (mostly due to bad terminations) can cause all sorts of network strangeness and intermittencies.

I would like to know a bit more about what "not working" means in relation to a computer.  What observations are made to reach this conclusion?
- ipconfig
- ping
- trace route
- name service
- internet access
- LAN access
Which of these end up with abnormal results?
My first guess would be the computers with the bad/no connectivity are booting up and asking for addresses before the DHCP server is ready. Did you try running the network diagnostics on each computer to see if it restores the connection? (or manually disable the network adapter, wait a few seconds then enable the network adapter again... that's usually the first thing the network diagnostics wizard does anyway.)

As an electrician, I see 3 solutions...  :)

1) install a generator + transfer switch to power the computers when the grid drops.
2) size the UPS's so they don't run down and turn off during the longest outages.
3) have monitoring software do orderly shutdown of entire LAN if power is off for more than 5 minutes.

for 1), you could keep the same UPS units and delay the generator starting for a few minutes so it's not running unnecessarily.

Featured Post

Problems using Powershell and Active Directory?

Managing Active Directory does not always have to be complicated.  If you are spending more time trying instead of doing, then it's time to look at something else. For nearly 20 years, AD admins around the world have used one tool for day-to-day AD management: Hyena. Discover why

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