This post is more than 5 years old
29 Posts
0
29454
To Denial,,,
Hello Daniel, I think you are in a whirl of your business.
If you are Okay, can you help me about network setting?
It's a little bit long story.
At first, I'll show you my topology.
As you can see the topology, there are 3 switches ; Dell 8024F, Dell 7024P, Cisco SG300.
Actually, I originally set up the machine to L3 mode except for Cisco SG300 devices (L2).
So I set these switches like followings.
1. Routing protocol at 7024, 8024 : OSPF (L3)
2. Dualization between Dell 7024 : VRRP (L3)
3. Vlan 1 ip setting at SG300 : L2
That is, above 7024 switches, L3 mode, other side is L2 mode.
Did you get it what is going on..?
If you understand until this, I'll mention continuously.
After setting, everything is Okay.
That is, the whole devices communicated each other except for that I indicated "My company's Display Extender" which is used for TCP, UDP in video at the below of left side.
When I used "TCP" at Display Extender in video, it was going alright.
But, I used "UDP", It was not going alright.
To sum up this, like below picture.
Do you think this situation might be characteristic for own our device?
For that reason, I changed from L3 to L2.
In this point, only thing that I set was vlan IP assignment.
Until this also, it was going alright.
A few days later, some issues occurred.
1. When I added 8ea x 7024P, 32 x SG300 below 8024F, the whole network connected 7024P was disconnected suddenly.
=> We set SG300 only VLAN assignment.. so I supposed it must be going well..
Perhaps would you think the traffic is cause the disconnection due to lots of devices?
2. When I changed "Display Extender" device to another Extender, it was not communicated properly even though another Extender's IP was same as previous Extender.
=> In theory, It was out of common-sense. I don't understand what is this..
★3. Related to 8024F, I'll let you know what is the problem.
First, I pick out the ①cable , communication (OK)
Second, I put the cable original status.
Third, I pick out the ②cable, communication (Failed)
How can this situation happen?;; In the theory, I don't understand, maybe you too.
I really appreciate all you efforts.
Please contact me..
Thank you expert.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
1
October 28th, 2015 06:00
Currently with all of the devices set to the same priority the switches will elect the root switch based off MAC addresses. This creates a scenario where the root switch could be one of your SG300 switches, and that is not ideal. I would use the show commands to gather information and map out the current root switch, which ports are designated, root, and discarding. Knowing this information can help explain why you are seeing communication on some links but not others.
The spanning tree changes can cause some disruption in the network as topology changes occur. Be sure to plan out what changes are going to be made and do the changes during planned non peak hours.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
1
October 27th, 2015 08:00
It sounds like spanning tree could be playing a factor in some of the behavior you are seeing. I would run the following command on the switches: # show spanning-tree. This will provide you information on the spanning tree mode, priority, and port status. For even more information you can add "detail" to the end of the command: # show spanning-tree detail.
Check and make sure all switches are running the same spanning tree mode. The default for most switches is RSTP. You can change modes with the command: #spanning-tree mode rstp
Ensure that one of your core switches is the root switch. The default priority is 32768, by setting a lower priority you can ensure that the desired switch becomes root. # spanning-tree priority 4096
Desired root switch = 4096
second 8024f = 8192
7024 switches = 28672
SG300 switches = 32768
Ensure that your switch to switch connections do not have portfast enabled. Poirtfast should be enabled on edge ports that connect to the workstations.
All of the switches should default to a pathcost of long. When you issue the show spanning tree command the cost for each interface should be a higher number.
Name State Prio.Nbr Cost Sts Role Restricted
--------- -------- --------- --------- ---- ----- ----------
Gi1/0/3 Enabled 128.3 200000 FWD Desg No
If you see a low number, it means short mode is being used.
Name State Prio.Nbr Cost Sts Role Restricted
--------- -------- --------- --------- ---- ----- ----------
Gi1/0/3 Enabled 128.3 100 FWD Desg No
Usually when an even happens on the network, such as part of the networking going down, there will be some messages logged. The messages can help isolate the cause of the event. You can use the command # show logging, to see what has been recorded.
I am not sure about the display extender. Do you have a model number for the display extender? I can help do some research on it.
ksk79174766
29 Posts
0
October 27th, 2015 19:00
Thank you for your kindly answer.
About display extender, I can't tell you specifically..sorry!..
By the way, I think I don't focus on STP because the switches implement STP by default.
So only I have done is assigning the IP in the VLAN 1.
As your opinion, should I configure STP passively?
Regards,
ksk79174766
29 Posts
0
October 28th, 2015 03:00
I want to asl you some questions.
1. As you mentioned,
should I set the root switch at core?
For example, if I set the root switch other switches except for core switch, how will this network be going on?
2. Then if I set the whole network devices based on your advice, may does this communication go well?..
Actually the most worry that I think is about disconnection all of devices.
I heard that kind of situation..
Regard,