![]() |
|
#1
|
|||||
|
|||||
|
What do I need?
We are having issues that we are pretty sure related to the network infrastructure, such as random losses of connections over the entire network (passing through several switches to reach the DC) amongst other things, so we want someone to come in and work out exactly what is causing these random issues on the network, whether it be switches, the cabling, the fact we have cat5 as opposed to cat6. A problem with the fiber network at some point that links it all together. It literally could be anything but from a front of house point of view, replacing machines, moving ports on switches, changing network cards and so on makes no difference. Is this a network analyst we need? or is there some other role that is more suited? I think it would be a network analyst wouldn't it? and how much do these usually charge? We are looking for someone to come in, work out what exactly is the issue if possible, so that we can decide on whether it warrants spending money on fixing. Passed: 271,272,680,682,621,640,642, PRINCE2 F, ITILv3 F Or for easier reading, the below: MCP, MCDST, MCITP:EST, MCITP:EST7, MCTS WIP for 2010: 290,291,293,294,298,351,643,646,647,686 Or for easier reading, the below: MCSA:S, MCSE:S, MCITP:EA, MCITP:SA and MCITP:EDA No I don't expect to get anywhere near that, but a target is a target |
|
#2
|
|||||
|
|||||
|
i guess that depends on the infrastructure really.
Has the building had all the cabling and patch panels tested and certified? |
|
#3
|
|||||
|
|||||
|
Send me a copy of your switches configs (Minus any sensitive stuff) and I will have a look at them.
Network diagrams would also be very nice too Cheers Joe the worse thing you can do is to do nothing, so do something, anything, everything!!
|
|
#4
|
||||
|
||||
|
Any person with the experience, i wouldn't tie it down to "network analyst" as there will be many other people. Your best bet is going to a IT support company with network expertise on the payroll.
Out of interest, what are these outages? do they last long enough for you to do pings to different areas of the network to find out where it stop working? |
|
#5
|
||||||||||||
|
||||||||||||
|
Quote:
Quote:
For the most part it seems to run perfectly adequately, however we they really need someone who specialises in networks to come have a poke around for a day or so and give some advice out. Quote:
Passed: 271,272,680,682,621,640,642, PRINCE2 F, ITILv3 F Or for easier reading, the below: MCP, MCDST, MCITP:EST, MCITP:EST7, MCTS WIP for 2010: 290,291,293,294,298,351,643,646,647,686 Or for easier reading, the below: MCSA:S, MCSE:S, MCITP:EA, MCITP:SA and MCITP:EDA No I don't expect to get anywhere near that, but a target is a target
|
|
#6
|
||||||
|
||||||
|
Quote:
Firstly I would look at spanning tree. Have you got all the access ports manually set to "spanning-tree portfast" = This will reduce the spanning-tree timers on access ports Have you got all the access ports manually set to "spanning-tree bpduguard enable", shut down ports in the event that a "rogue switch is added to the network Are you using STP 802.1D (Cisco know it as PVST) or 802.1W (Cisco refer to it as Rapid-PVST) Have you disabled "DTP" from all the links between switches and made every link a static trunk? Do you have UDLD enabled between switches? Are all the uplinks full duplex? Have you got all of the switches set to "vtp mode transparent", stops the buggers from wiping each other out. Have you idntified if any of the ports are going into "err-disable", I think top of my head the command to view all the err-disabled ports is "show interface status" Could not suggest anything else unless I saw the switch configs. Cheers Joe the worse thing you can do is to do nothing, so do something, anything, everything!!
|
|
#7
|
||||
|
||||
|
"show int status err"
|
|
#8
|
|||||
|
|||||
|
the worse thing you can do is to do nothing, so do something, anything, everything!!
|
|
#9
|
||||
|
||||
|
|
|
#10
|
||||
|
||||
|
How is the cabling? Is it near any power cable? are there any times during the day that this occurs (say between 8.30 and 10, 12-2pm etc) or is it completely random (ie perhaps large amounts of data going across the network that's saturating it (people pulling PST\DB's across the network).
What do the logs on the servers say? Are they dropping the NICs or is it switches? I would perhaps look at running Wireshark just to see if you can see something on the network that's perhaps causing it. I would also what you can ping when the connections drop, can you ping between switches or just on the same switch, is it the uplinks that's causing you issues? CNA | CNE | CCNA | MCP | MCP+I | MCSE NT4 | MCSA 2003 | Security+ | MCSA:S 2003 | MCSE 2003 | MCSE:S 2003 | ITIL Foundation v2 | MCTS:SCCM 2007 | MBCS | MCTS:Win 7 | MCITP:EDA7 | MCITP:SA | MCITP:EA | MCTS:Hyper-V Disclaimer: The views of SimonD are just that, his views. He doesn't suggest or recommend that others live their lives by following his example. |
|
#11
|
||||||
|
||||||
|
Quote:
Passed: 271,272,680,682,621,640,642, PRINCE2 F, ITILv3 F Or for easier reading, the below: MCP, MCDST, MCITP:EST, MCITP:EST7, MCTS WIP for 2010: 290,291,293,294,298,351,643,646,647,686 Or for easier reading, the below: MCSA:S, MCSE:S, MCITP:EA, MCITP:SA and MCITP:EDA No I don't expect to get anywhere near that, but a target is a target |
|
#12
|
|||||
|
|||||
|
BosonMichael MCSE+I, MCSE: Security, MCSE: Messaging, MCDST, MCDBA, MCTS, OCP, CCNP, CCDP, CCNA Security, CCNA Voice, CNE, SCSA, Security+, Linux+, Server+, Network+, A+ Served proudly, US Army, 98C Intelligence Analyst, '89-'92 Everyone else is posting their blog... guess I will too! |
|
#13
|
|||||
|
|||||
|
Hi Daniel
I have a suspicion that your problem may be due to "Errdisable linkflap" causing the ports to shutdown rather than UDLD detecting unidirectional links. Get the syslog up and have a gander. Cheers Joe the worse thing you can do is to do nothing, so do something, anything, everything!!
|
![]() |
|
||||||
| Bookmarks |
| Thread Tools | |
| Display Modes | |
|
|