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: 219
  • Last Modified:

Connection lost to some servers

Dear Experts,

I installed two switches ; UP switch and Down switch ; as shown on the diagram attached, I configured both switches with VRRP between the two switches , these switches are connected to two core switches ; core 1 and core 2; core 2 should be the backup of core 1.
Both UP and Down switches are connected to a rack with many servers inside, these servers have IP's in networks 10.2.17.0 , and the gateway of servers is 10.2.17.1 which is configured on UP switch for vlan 1.
When someone try to connect ping to server on the rack some servers loose connection and the extended ping oscillates between reply and request timed out.
I attached the diagram and configuration of both switches, I need also to know if VRRP is configured correctly because it is the first time I'm doing it in production network?

Thanks In Advance,
Diagram.pdf
Down.docx
UP.docx
0
oamal2001
Asked:
oamal2001
  • 5
  • 4
2 Solutions
 
aleghartCommented:
Why are you connecting only your edge switches?  Why don't you connect your cores together, then redundant path from each edge switch to each core?

From this diagram (connections) it looks like your "core" is something like an internet or WAN router, not a core.
0
 
oamal2001Author Commented:
Hi,

Sorry I forgot the redundant link between the core switches, I modified the diagram and attached it.
I need to add that when I try to ping from down switch to networks 10.2.17.0, 10.2.18.0 and 10.2.20.0 I cannot ping.

Thanks,
Diagram.pdf
0
 
aleghartCommented:
If your cores are connected, then your edge switches do not need to be connected to each other.  They connect only to the cores.

I guess I'm asking, why are your running VRRP/HSRP on edge switches instead of the cores?

Cross-connect Core1 and Core2 with VRRP/HSRP
Connect(trunk) SwDown to Core1.
Connect(trunk) SwDown to Core2.
Connect(trunk) SwUp to Core1.
Connect(trunk) SwUp to Core2.
Do not connect SwDown to SwUp.

Maybe I'm misreading something.  I'm no architect.
0
VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

 
oamal2001Author Commented:
The VRRP between the two down and UP switches are required because they will handle a critical system, and all servers will be connected to both switches, I'm responsible of these switches , I cannot access the core switches.
I need to know if this design can be implemented to work fine or not.

Thanks,
0
 
aleghartCommented:
Yes.  You're collapsed differently.  Your design is not collapsing core+distribution (my assumption, sorry...didn't realise that you can't control the cores).  Your design is collapsing access+distribution.  This collapsed layer is your boundary between L2 & L3.  This boundary is the place to have VRRP/HSRP.

Your addressing is a little confusing.  
Your SwDN doesn't have an IPv4 address for vlan 1.  SwUP is holding the .1 address.  The virtual IP is .3
My numbering would be something like this:

interface vlan 1
ip address 10.1.1.1 (virtual IP shared by both switches)
ip address 10.1.1.2 (SwUP)
ip address 10.1.1.3 (SwDN)

interface vlan 2
ip address 10.1.2.1 (virtual IP shared by both switches)
ip address 10.1.2.2 (SwUP)
ip address 10.1.2.3 (SwDN)

interface vlan 3
ip address 10.1.3.1 (virtual IP shared by both switches)
ip address 10.1.3.2 (SwUP)
ip address 10.1.3.3 (SwDN)

This makes it easier for all vlans...the gateway address for the subnet is always the .1 address.

I'm not sure if it's inadvertent, but your current vlan addressing are publicly routable address spaces.
12.0.0.1 belongs to AT&T Services
20.0.0.1 belongs to Computer Sciences Corp
Any misconfiguration or routing problem, and traffic will go to default route (0.0.0.0) which would be expected to jump to your core and out to the internet.

I fight that problem all the time...somebody "designed" with IP addresses that are _not_ private, non-routable IP addresses.  Anything that isn't handled explicitly by a created route will jump to the nearest internet connection.

I hope this makes sense.
0
 
oamal2001Author Commented:
What about address 10.2.17.1, I need to make this address the gateway for all servers which connected to both switches.

Thanks,
0
 
aleghartCommented:
10.0.0.0/8 (anything that starts with "10.") is private address space.
0
 
oamal2001Author Commented:
Kindly, How to configure 10.2.17.1 on the switches to be the default gateway connected to them?
Thanks,
0
 
oamal2001Author Commented:
HI ,

I configured one of the switches to disable spanning tree with no spanning tree, does this affect?

Thanks,
0
 
Johneil1Commented:
"Kindly, How to configure 10.2.17.1 on the switches to be the default gateway connected to them?
Thanks,"

--If you would like for the IP above to be a DG for your servers....then i would not use the management VLAN1 and create a new SVI (e.g. VLAN 10) with the interface IP of 10.2.17.1 and then place servers in that vlan.

This LINK may help you. with quoted text above. Let me know what you need for the other stuff.
0

Featured Post

Technology Partners: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

  • 5
  • 4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now