Meraki

thomasthomsen

Fingers crossed if I can now get Smart Port Automation to configure a trunk port correctly to an AP. Not on the bug fix list, but one can hope. Will test later.

thomasthomsen

By doing as you did "unhide" the SSID ? Wireless -> SSID availability -> Visibility = Advertise this SSID publicly. Or perhaps I do not fully understand your question ?

thomasthomsen

I got the answer through other channels. (I dont think Im allowed to share here). But Ill just mark @Inderdeep's answer as the solution. 🙂

thomasthomsen

And this "new image train" is 19.x ? Or something completely new (Since it seems that the new MX650 is based of a Cat 8K - in theory it could be IOS right 🙂 ).

thomasthomsen

What exactly is the wireless NIC chipset ? model ? Does it "see" your SSIDs on other of your APs (if you have other models of AP). Historically I think the only times I have seen clients not being able to "see" a SSID has been if the drivers where old, and they did not support newer technologies like 802.11ac / ax and so on. You can still disable for example .ax in your RF profile, but your MR33 is not a .ax AP, so I dont think that is the issue. I have also seen some clients having issues with hidden SSID (and hidden SSID is "NOT a securityfeature(tm)" ) ... so start sending those beacons containing the name if you have disabled that 🙂 ) Also, a little related, the band steering feature can cause problems, basically for the same reason. Just my thoughts. /Thomas

thomasthomsen

If you just ping a random wireless client, that is just "sitting there", it might show you a graph that looks like yours, because a lot of wireless clients do power save on their wireless NIC. This means it signals the wireless network that it is now going to sleep, the AP buffers the traffic, and advertises the client (traditionally in beacons) that there is traffic for it and it should wake up. The clients NIC then polls the AP for the buffered traffic. This of course is something that happens very quickly, but it will show up in fx. continuous ping graphs. Sometimes, but not always, depending on the client, OS, driver and so on, the wireless NIC might not go to sleep while connected to power. If you experience trouble while just doing normal operation (work) on a client, then, as many has suggested, you start troubleshooting at layer 1 and do a wireless sitesurvey, to ensure that signal strength, interference, noise and so on, are ok, so that the client can get the signal it needs, in order to perform well for your applications. /Thomas

thomasthomsen

Thanks But will it be implemented in ALL MX (both virtual and hardware) ? .. just the AWS vMX ? or all vMX ?

thomasthomsen

If you are the admin on both orgs, you can move licenses around yourself today (or, you can get support to help you). You could also have one org, and give different access to each user, so they only have admin rights to the networks that they use ?

thomasthomsen

Do we know from what version of the MX software that vMX will utilise VPP ? Or is this only for AWS vMX-XL ? I mean would all "clouds" and versions of vMX not benefit from VPP ? Cheers Thomas

thomasthomsen

Very interesting - Lets see what happens.

thomasthomsen

And it also seems fine for me on MX85.

thomasthomsen

I was wondering the same thing. So ... "during" the upgrade the MX can misclassify traffic ??? .. until the upgrade is done at it reboots ? Is that what it is saying ?

thomasthomsen

I was also wondering about Split DNS ... but perhaps its for Anyconnect ? - Or do you mean the API support and not the feature as documentation ? /Thomas

thomasthomsen

Some Apple documents that might be of interest for some. https://support.apple.com/en-jo/guide/deployment/dep3b0448c58/1/web/1.0 <- See the AutoJoin section, for some "dbm" numbers. https://support.apple.com/en-gb/guide/deployment/dep98f116c0f/web <- See the "Selection criteria for band, network and roam candidates" for some "trigger" values, how much better the new signal must be, and preference in wide channel, and "Wi-Fi generation" when roaming.

thomasthomsen

It also only seemed to do it for a "short while" , and I could see all these clients on the summary page, with all the public IP addresses (but , as far as I can remember), with the MAC of the MX on all. But apparently they where all counted as a new client. It stopped, but I dont know how long it took, and apparently the dashboard kinda remembers that, its no biggie, but it now (until the time experies on that data) an "eye sore" or whatever its called 🙂

thomasthomsen

Playing around with C9200L in my lab in the "IOS-XE Cloud native" mode - "60% of the Time, It Works Every Time". (And I still dont quite know what marketing would call this, "persona" like on the APs 🙂 or are we going with monitored / managed ?) Jokes aside ... Something fun seems to happen. I think the switch starts to count ALL IP addresses in flows as clients, for a while. At least Im quite certain that I have not had over 4000 different clients in the lab network across a weeks time. Have anyone else seen this ? /Thomas

thomasthomsen

Upgrading right now 🙂

thomasthomsen

Some clients (In my experience especially Apple computers, and I think it is even in their documentation) REALLY prefer wider channels, so if you are running "auto" on channel-width, and one AP is at fx. 40Mhz and the other one is at 20Mhz , then the client might prefer the AP with the 40Mhz channel, even though it is far away. I always recommend "static" channel-width, so 20,40,80 depending on your local regulatory domain, and whats feasible.

thomasthomsen · ‎Oct 19 2024

He he .. I didnt know you would be celebrated this way. Thanks all. The only thing that really "bugs" me is that ratio of posts / kudos / solutions.... , apparently Im not a "solution oriented" person 🙂 - who knew 🙂 (And @cmr and @KarstenI I almost always feel motivated 🙂 )

thomasthomsen · ‎Oct 17 2024

I never noticed that in the documentation. Oh well .. not a bug ... but a "feature" then ... 🙂 Lol and also the LAN interfaces I see ... what kind of horror is this : State (enabled/disabled) of any LAN interface

thomasthomsen · ‎Oct 17 2024

That is quite strange (a disappointment even ?) because the 9800 WLC uses the native Nextunnel for monitoring ever since 17.12.3. But I can wait, and use the old "app" for putting switches into monitoring mode. I do have some customers that dont like that way , at all, and they are just waiting for the rest of the Catalyst switches to work like the C9800.

thomasthomsen · ‎Oct 15 2024

With the introduction of 17.15.1.9 (or 17.15.1m1 as the IOS-XE version is called "on box"), we got the tunnel native in IOS-XE (like on the 9800 WLCs). But this, so far on the switches, seems to only work for being used for management, not monitoring. How will monitoring be accomplished using the "cloud native" method ? I know the command : "test platform software config mode managed" will start the migration to Meraki managed mode (after you have started the new native tunnel, and the switch has got the Meraki serial and so on), or adding your switch from inventory to a network should start the same process. I also, kinda, noticed that there is a "sister" command : "test platform software config mode monitored" <- Or something close to it (its from memory 🙂 ). Is this the command we should use for monitored mode in the future ? Can we still then add the switch to the inventory, and add it do a network without the process for migration to managed mode starting ? Perhaps Im just to impatient, but I kinda, really dont like the method currently being used with TLS interfaces and so on, when this new, seemingly, much more modern solution is now there 🙂

thomasthomsen · ‎Oct 15 2024

5 Ghz non-wifi noise (throughout a building), all the way from channel 104 to above 165. With a clear center channel around 149-153 ish. .... Now this is a real wireless nightmare.

thomasthomsen · ‎Oct 15 2024

I kinda did the same thing in the home lab 🙂

thomasthomsen · ‎Oct 14 2024

I had a weird thing happen to me as well. "Shut / No shut" WAN 2 on my MX85, and it took down the network for, what felt like, a couple of minutes. 😕 WAN2 is connected to a MG, and WAN1 is of course just the normal internet connection, and WAN1 of course has the priority (just to be clear 🙂 )

About thomasthomsen

thomasthomsen

Thomas Obbekær Thomsen

Community Record

Badges