MS-390, when is it ready for action?

Darren8
Here to help

Re: MS-390, when is it ready for action?

I was very tempted but didn't want to risk it.

 

Support also said unless they tell you to for a specific reason you shouldn't have to (ie stacking should work with the stable release).

 

cmr
Kind of a big deal
Kind of a big deal

Re: MS-390, when is it ready for action?

@Darren8 with the MS390, the versions after 12.28 bring you fixes about what client is reported where, PSU info, PoE info, routing fixes, crash fixes, the ability to change settings and not need to reboot the switch and more.

 

The only known downside of 14.8 that applies to the 390 is that the stack must be no more than 3, but you are okay with that.

MichelRueger
Building a reputation

Re: MS-390, when is it ready for action?

I just do Testing with MS14.8 Beta on my MS390 and yes there are some updates Like Power supply information but still a lot of things not working. Client Information is now OK. But you can not have more than 3 Switches in a Stack with Beta 14.8!!

 

Kabel Test is not Working.

Ping from the Switch goes up to more than 400ms and sometimes even up to 2 seconds!

Bildschirmfoto 2020-12-06 um 20.26.02.png

 

hope to be able to post more updates soon.

Darren8
Here to help

Re: MS-390, when is it ready for action?

Got an outage window and back onsite and basically almost 5 hours later, 3 Meraki support engineers and still no data stack.

 

I followed the provided instructions which failed at step 1. Found a way around it and continued. Waited 3.5 hours after the final step for the stack to form and then called it.

 

Went with 12.32 this time.

 

Not a happy camper 🙄

whistleblower
Getting noticed

Re: MS-390, when is it ready for action?

hi @All

do you still think the same way today? how is the experience with the MS390 series as of today? I’m also considering a MS425 flex stack in a new deployment and use the MS390‘s as physical stacks as access switches! One reason for me is that I was told that the 390 would be the best for future use - e.g. of the hardware specifications, Adaptive Policy feature etc.

Maybe someone knows if SGT would also be supported on the other switch models like MS425, 350 and 250?

cmr
Kind of a big deal
Kind of a big deal

Re: MS-390, when is it ready for action?

@whistleblower I'd use 355Xs as an access stack, in fact I have replaced a stack on 3850s with 355Xs and the Cu 2.5/5/10GbE is a great benefit for access.  As they have QSFP+ ports we are going to use those to link to our core at that site.

Bruce
Kind of a big deal

Re: MS-390, when is it ready for action?

I’m still in two minds about the MS390. From a futures perspective, yes it will be the best switch. Should you deploy it now? Hmmmmm... if you just want switches doing nothing too special in small stacks (three or less), and you’re happy for the odd hiccup and to run the MS14 ‘beta’ code then I reckon they’re probably deployable.

 

As for SGTs, I doubt very much that the traditional Meraki switches will ever support them. The Cisco Catalysts and the MS390 that support SGTs have the UADP ASIC in them so that they can process the additional payload in the Layer 2 header - its unlikely the traditional switches will ever get this capability as their ASICs most likely can’t do it, and doing it on the CPU is impractical.

 

I expect, but this is pure speculation, that potentially once the MS390 has reached a certain level of maturity and stability then Meraki may bring a small number of other Catalyst hardware over to the MS range, like the Catalyst 9500 to expand/extend the MS400 series. Under the hood the Catalyst 9000 switches are all similar from my understanding, so once one is done the others should be significantly easier (and probably not such a mess as the MS390 was).

GIdenJoe
Kind of a big deal

Re: MS-390, when is it ready for action?

I strongly agree with @Bruce here.

 

The catalyst switches themselves are very stable but the integration with Meraki software is going a bit bumpy.

Once they got the hang of it, porting other members of the catalyst portfolio to the newer MS line will be less troublesome.  All catalyst switches use the same kind of UADP chip in one generation (except the 9200's) and the software behaves the same over the entire family.

 

The traditional MS line is more stable but has way fewer supported features than a catalyst switch,

 

Smaller customer that exclusively route traffic over the upstream appliance albeit an MX or a 3rd party device will be just fine with the current MS family.

 

However if those customers grow and go to a routed distribution layer/core then segmentation becomes more important and that is where the trustsec/adaptive policy value lies.

MichelRueger
Building a reputation

Re: MS-390, when is it ready for action?

I am in the same Status. As the Firmware with solves a lot of Problems is still Beta I do not recoment to buy it now. Or only if you need Multigigabit Ports. One of the biggest problem still is that if you have a stack with ms390 and you are updating software the full stack goes down!! and not just for 3 Minutes. Last update of a stack with 4 MS390 took more that 35 Minutes to get fully operational again.

 

I also See that Meraki is using more Cisco Nativ switches in future. But if you see how it is implemented at the moment. this will never work stable. if you open the MS390 Switch you find the Serial Console Port. Connect a Console Cable to it and power on the switch. You are able to see that the Switch is powering up a Cisco nativ IOS and then starting a virtual stack for the Meraki implementation.

whistleblower
Getting noticed

Re: MS-390, when is it ready for action?

thanks  @Bruce @GIdenJoe 

 


@Bruce wrote:

I’m still in two minds about the MS390. From a futures perspective, yes it will be the best switch. Should you deploy it now? Hmmmmm... if you just want switches doing nothing too special in small stacks (three or less), and you’re happy for the odd hiccup and to run the MS14 ‘beta’ code then I reckon they’re probably deployable.


...and that`s the point, for now the switches will be used for wired-access /w 802.1x and to connect APs maybe via mGig... for this to achive the 355 Series would be fine! But when it comes for future use and one point out of that would be in my opinion Adpative Policy/SGT than I`d have to go with the MS390...

 

The thing is... when I use the MS425 in the core and they won`t support the Adaptive Policy feature neither in the future as well - the design is`nt as good for future proof! So I`d to use the MS390 (/w all differences e.g. like flexible-stacking compared to the MS425...) or mix Meraki MS-Switches with Cisco Enterprise C9K devices - which I honestly would like to avoid! 😕

PhilipDAth
Kind of a big deal

Re: MS-390, when is it ready for action?

>when I use the MS425 in the core and they won`t support the Adaptive Policy feature neithe

 

If you buy the MS390 are you going to buy the "Advanced" licence as well to enabled Adaptive Policy?

whistleblower
Getting noticed

Re: MS-390, when is it ready for action?

hi @PhilipDAth 

 

yes, the plan was to buy the license to use the Adaptive Policy!

 

BTW - I‘ve attended the meraki PVT in the last week and the SE‘s advised that currently the way to go in using Adaptive Policy will be to work with C9K in the core! the questions - if the classic MS Series will also be adopted to use the same future proofed features like the MS390 offer today... where’nt answered very clearly to me 😕

 

In my case now, if the customer will decide to use micro-segmentation, I‘ll change all over to Cisco C9K switches for core/distribution and access layer! If not, than I probably choose the MS355 as the access switches because of the better stability...

Bruce
Kind of a big deal

Re: MS-390, when is it ready for action?

@whistleblower, it’s highly unlikely that the classic MS switches will ever be able to do Adaptive Policy. AP uses a Scalable Group Tag (SGT) which is inserted into the Ethernet header, and since virtually every switch uses ASICs to process the Ethernet header, the ASIC needs to ‘know’ about it. Unfortunately the ASICs in the classic MS switches aren’t reprogramable.

 

The MS390 has Cisco’s UADP ASIC in it, which is the same as in the Catalyst 9000 series, hence why they all support SGT and why they can all play nicely together. The UADP ASIC also supports a level of reprogamability too, which is why Cisco rave about it so much - means they can add new capabilities to the ASIC without throwing the switch out.

misterguitar
Here to help

Re: MS-390, when is it ready for action?

HA! I'd be happy just to get a #$%^& MS390 to form a stack correctly without missing stack ports in the dashboard. I'm not even worried about advanced features on the 390. I just want the basic ones to work!

BazzaP
Conversationalist

Re: MS-390, when is it ready for action?

We had the dashboard dropping also, when you looked on the 9500 core it was complaining about low power on the links to the core. Just upgraded to 14.21 so far all good, been just under a week up now.

watdee
Here to help

Re: MS-390, when is it ready for action?

I like the size and the fact that it can have 8 10gb connections, plus I need the removable brackets to get it in my rack.  I'd like to use a pair of them for my server rack but worried about the boot time, bugs, and missing features.  Doesn't seem like this switch will ever be ready for prime time.  They have the MS390 on the switching home page, is this really the future of Meraki?

BazzaP
Conversationalist

Re: MS-390, when is it ready for action?

Hi You need to be on the latest firmware, currently 14.24. This has improved boot times significatly. There are still issues with these switches but they are getting better. They do have their own minds when it comes to stacking and aggrigating ports. Always best to get them online first before you do anything with them, this breaks from the usual pre-staging that you would normally do with a switch network.

misterguitar
Here to help

Re: MS-390, when is it ready for action?

It would depend on the criticality of what is going on the switches. Currently they are not stable enough for use in situations with servers or large traffic loads for us. They are relegated to being access layer switches for the time being. Frajnly we have been trying to figur eout what th eplan is here with regards to Cisco and these switches. It is pretty clear to me that the switch is running Catalyst IOS XE natively and the Meraki software is running in an IOS XE VM to provide the "Meraki" interface. Basically with the Meraki software VM acting as a "thunking" layer. I have verified this with Meraki support folks who had access to the catalyst stuff going on underneath for debugging purposes for some of our open tickets. Issues we had included problems with Stacking, stack members randomly reordering themselves, stacks requiring reboots for layer 3 SVI changes to take effect, issues with QoS, the 1000 vlan limitation issue, and issues with STP compatibility between catalyst IOS switches and the meraki switches. When doing forklift network upgrades for campuses, we have been having to build entire parallel networks back to the firewall to isolate the STP from the old catalyst STP or you get all kinds of weird stacking problems.

 

I come from a CCNP background and I have scoffed at our management's attitude that somehow this stuff is the future of networking. We deal with a lot of clients who have high turnover low skill set engineers due to low wages and living in rural areas. (Think government workers.) So they bought the Kool-Aid on this we can avoid fixing our HR problems and paying real salaries by getting this "Anybody off the street can build a GUI network" kit. The actual old Meraki hardware has been relatively trouble free. But having to debug a buggy beta product foisted on the masses with no actual debugging tools becasue they are hidden has been a HUGE time suck. Our early projects were regularly over budget and late due to trying to debug these things. Once you know the tricks like:

 

1. Isolate the STP instances from any Catalyst STP

2. Upon brining up an MS390, set all ports to access and VLAN 1 to avoid the 1000 vlan problem when connecting to switches that pass all vlans rather than a subset on trunks. 

3. After making MS390 L3 changes reboot the stack to make sure things take effect.

4. Bring a magazine and some coffee to pass time while the switches are booting up. Bring a book if you are doing things that require multiple reboots like firmware updates.

 

My speculation having worked for government and large outfits is a very large company was working with cisco on a very large business deal. This deal probably stipulated port requirements, power requirements, feature requirements, or redundancy requirements that existing Merki hardware could not provide. So to not lose the contract, Cisco got a couple of engineers to hack together this kludge so they could be a merki solution but meet the hardware requirements. And once it was done they started selling this to the general public. But yeah the view here is this thing is still Beta.

Get notified when there are additional replies to this discussion.
Welcome to the Meraki Community!
To start contributing, simply sign in with your Cisco account. If you don't yet have a Cisco account, you can sign up.
Labels