1

Hello,

We have 5 stacks of Cisco SG500 52port PoE switches.
One of the stacks has developed an annoying issue where at random intervals the entire stack will soft reboot. As this is the stack that all the other stacks are connected to, it causes the entire building to lose internet access for a couple of mins.

Our managed provider of the switches has said that they are EoL and Cisco will not help them get to the bottom of the problem, the only suggestion is to replace the units in that stack with Meraki equivalents.

Can anyone see there being a problem in replacing one stack with new Meraki switches and leaving the other stacks in place, or should I look to replace all of the stacks with new switches?

Thanks, Stephen

5 Spice ups
2

Unless/until the others begin rebooting (provided they’re working as expected) is to leave them as-is but plan/budget to replace as needed. They may develop the same trouble and they may not. Either way if they’re EOL you might be considering eventual replacement.

3

… then there’s the upgrade cost. meraki (especially in the uk), are exponentially more expensive compared to cisco (same company i know), but that’s just how it is.

the other thing that you need to factor in is the cost for the licencing, without it, and cloud access you have a managed hub that you can’t manage.

it might be worth asking your service provider the costings for refurbed units.

not sure if that’s an option, that would be available to you?

2 Spice ups
4

Thank you.
The other stacks are currently working correctly.

My main concern (as I don’t know how stacking works) is that if we removed a stack, does that affect the other stacks?

5

Good question. And one I think that someone who has more experience than I with Cisco/Meraki should answer. Theoretically, if configured 1:1 (VLANs, ACLs, etc.) all should play well together.

1 Spice up
6

I haven’t seen pricing for Meraki in a while, nearly 10 years. I used them from 2014-2017.

If it were me, I would replace your existing with Cisco Catalyst 9200 or 9300 series switches with the stacking option.

7

There’s no technical issues. The issues would be around costs for hardware and licensing, and operational expenses managing two different types of switches. That said, Cisco Meraki are cloud managed and should be pretty easy, and it’s your network management company that deals with the operational issues.

The Meraki switches will give you way more information about what’s happening on the network, applications used, traffic, etc. That’s a major upgrade if you don’t get that from your firewall. On the downside is that they’re really expensive switches.

Personally I buy used Cisco Catalyst, which have been very reliable over the decades. If you can afford Meraki switches, they’re pretty awesome.

1 Spice up
8

If you do not want to use Meraki the direct replacement of your current product is the CBS350-48P-4X. the CBS350 range support true stacking like your current switches.
If you do not need ‘stacking’ and would be happy just uplinking/daisy-chaining switches the CBS220 range (cheaper) would also be fine.
The Catalyst 1200 series are the lowest entry point full enterprise switch (no stacking).

Whatever new model is introduced make sure to configure the spanning tree and vlans + uplink ports correctly.

you may be able to solve the current stack issue also - how many in the stack? Try changing the master to another physical switch. what do the logs show as the reason for reload?

9

If Jurassic Park were built today, it would have Meraki switches.

10

Do you mean they are stacked ?
Usually replacement of stacked switches require same make & models or at least of the same brand ?

Do you have access to the logs to see why they are rebooting ?

  • some vulnerability ?
  • some bugs ?
  • excessive PoE power draw ?
  • looping or flooding etc ?

Then maybe the first step is to see how the stacking is done ? Using stacking modules, stacking cables or LCAP etc ?

11

Do you mean they are stacked ?
Usually replacement of stacked switches require same make & models or at least of the same brand ?

We have 5 stacks in total, there are 5 switches in the stack that has the issues. I am looking to replace all 5 switches in this stack, leaving 4 stacks (with about 4 or 5 switches in each of the other stacks)

Do you have access to the logs to see why they are rebooting ?

Unfortunately I don’t have access to the logs. Our provider have said they can’t see what would be causing it, but as they are EoL we should replace them, their suggestion was to go for Meraki.

Using stacking modules, stacking cables or LCAP etc ?

They are stacked using cables between the switches. (The black ones in the photo hopefully posted..)

switches
switches1512×2016 177 KB
12

I would think they are daisy chained from the pic…

  • stacking modules are usually at the back and quite unique if for more than 2 or 3 switches
  • core switch & distribution switch method is not used unless “core switch” have 4 or 5 such ports ?

Thus I could say you have 5 clusters of switches … each cluster have 5 switches daisy chained ? Stacking or stacks of switches usually mean they are “virtually made into 1 switch”

Can you get a copy of the logs from your provider ? Then go to the time just before the reboots as surely there is some line that says why the switches are rebooting or what user or account is allowing the reboot ?

Then one of the requirements of getting appliances might be that you may not want the “support” to be in subscription mode where if the support ends, you get no support at all (even to read logs) ? Else I would get maybe like HPe or Dell branded switches as they come with lifetime support (you may need to check how long the “lifetime” lasts, but many are way beyond 20 yrs but may require add-on support for like “1 to 1 exchange” and/or 4hr or NBD support).

13

Do you think it would be more risky to swap these 5 devices to Ubiquiti switches instead of Meraki?
A different provider has suggested Ubiquiti as they are less expensive.