Bruce123

We have already made the changes to fix the IOS9 issue, we actually tried resversing these on a test BSSID to see if they could have caused it, but it didn't help. We have raised a call with Meru (including logs of devices disconnecting/reconnecting), so if it is a known issue hopefully they may be able to assist. Thanks, Bruce. Thanks, Bruce.

Hi, We are experiencing intermittent dropped connections on our new HP Chromebook 11 G3 devices when connected to Meru AP320 APs. Typically, the Chromebook will stay connected for a while and will then lose connectivity (continuous dropped pings for a while). Sometimes the connection will be re-established automatically and othertimes it requires the user to disconnect and reconnect to the wireless (using the GUI). This is causing the Chromebooks to be unusable. We don't experience these issues when connected to AP332i or AP832e models of APs. Also, while the Chromebook is experiencing this, the wireless icon tray icon is still showing the connected icon and signal strength. The Chromebooks connect to the AP320 on .n and to the AP832e on .ac. ChromeOS versions 41-46 (haven't been able to try previous versions) Controller: 6.1-2-28 MC4200 An Acer Chromebook used as a benchmark is unaffected. Has anyone experienced this or similar issues? Many Thanks, Bruce.

Hi, We are experiencing intermittent dropped connections on our new HP Chromebook 11 G3 devices when connected to Meru AP320 APs. Typically, the Chromebook will stay connected for a while and will then lose connectivity (continuous dropped pings for a while). Sometimes the connection will be re-established automatically and othertimes it requires the user to disconnect and reconnect to the wireless (using the GUI). This is causing the Chromebooks to be unusable. We don't experience these issues when connected to AP332i or AP832e models of APs. Also, while the Chromebook is experiencing this, the wireless icon tray icon is still showing the connected icon and signal strength. The Chromebooks connect to the AP320 on .n and to the AP832e on .ac. ChromeOS versions 41-46 (haven't been able to try previous versions) Controller: 6.1-2-28 MC4200 An Acer Chromebook used as a benchmark is unaffected. Has anyone experienced this or similar issues? Many Thanks, Bruce.

Can I ask what date this was? Because when we contacted Meru, they apparently told us it wasn't an issue with Meru, but things could changed since. Thanks, Bruce.

Did they inform you what changes they made? Thanks, Bruce.

Do Meru still provide support for AP320 APs? Thanks, Bruce.

Has there been any updates on this? So far there seems to be two possible resolutions: Resolution 1: On each AP: radio parentbeacon radio0 never radio parentbeacon radio1 never (can be scripted to run on all at reboot so settings are not lost?) AND On the controller (ESS settings): Set SSID Broadcast on Set the SSID broadcast Preference for Vport is set to "Till association". Resolution 2: On the controller (ESS settings): Allow Multicast. Does anyone have any experience on which resolution is more effective? Does the issue lie with Apple or Meru (or both)? Is there a long-term fix due (either from Meru or Apple)? Does the IOS9.02 update fix it? We notice at the issue affects 320, but not 332i or AP832e Controller MC4200 6.1-2-28. Thanks, Bruce.

v10k. Further testing by ourselves and Websense support show that the issue doesn't relate to either SSL Interception or filtering (as these have been disabled - enabled SSL tunnelling - and the issue is exactly the same). This issue also occurs on our other two Websense v10K boxes. However, Websense support could not replicate the problem on in their lab. This is increasingly becoming issue as we are adopting Google cloud storage. Thanks, Bruce.

It sounds like a virus, or someone has the local admin password or a domain account with above basic rights. In my experience shutdowns that were initiated by shutdown.exe log an event in the event log (and would also give the username). Thanks, Bruce.

I passed CCNA on Monday. It was quite hard (not unexpectedly so) and I got stuck on one question early on (a practical) for about 40 minutes, which didn't leave me much time for the other 30 questions. Normally, my exam strategy is to skip any difficult/slow questions, with the aim of coming back to them later if I have time, but you can't do that with the CCNA exam as there is no "back" button. I would say the course/exam is around 50% Cisco CLI and 50% generic networking (up to layer 4), so even if you don't plan to configure Cisco gear the other 50% is still useful as it gives you the fundamentals. Also, bear in mind that some manufacturers try to emulate the Cisco CLI closely (such as Allied Telesys), so you're not just restricted to Cisco (and some aren't far off such as 3Com). However, the course/exam doesn't really deal with anything above layer 4, so if your interest is dealing with and understanding what goes on above these layers, there might be more appropriate options. Thanks, Bruce.

Hi all, We use Websense as our main proxy and have recently noticed some performance issues (particularly with uploads to Google drive and other services). SSL interception is enabled (which means that all SSL encrypted traffic is decrypted, scanned/filtered and re-encrypted). Post Snowden, an increasing proportion of sites are SSL enabled (e.g. practically all Google services it seems), so I am wondering what kind of impact this extra processing burden is having on our proxy (and our end user experience). Our Internet connection is 1Gbit. We recently measured file upload performance to Google drive and it came in at 20Mbit, which is acceptable for small files, but is not for large files (e.g. 1GB videos). Without Websense (or using Bloxx as a proxy, which doesn't intercept SSL) the throughput was a more respectable 120Mbit. We don't yet know if SSL interception is the culprit or it is caused by some other aspect of the way Websense is configured or functions (further testing is required). But my question is, has anyone else experienced poor upload throughput using Websense and what was the cause/solution? Has anyone noticed any proxy performance issues which relate to the processing of (the increasing amount of) SSL traffic (where interception is enabled)? Yes, I know not everyone agrees with SSL interception, but there is a strong case for it here in relation to Safeguarding. Thank you, Bruce.

Anyone else done the course and are preparing to take the exam? Thanks, Bruce.

Or that anyone with a debit card or paypal account and pay to use your broadband. Wouldn't it be better if you can take a share of this money? Bruce.

I had the same thought. I think the issue may be more to do with how to download the cert to the Android device and install it (through the browser). If you build an App which authenticates you by your username/password and then requests a cert from the server (which it generates on the fly), then in principle, it can handle the download and installation of the cert. If I do the same but by visiting a webpage instead, how will Android handle the download of the cert? Will it identify it as a certificate (through the extension) and allow you to install it into the certificate store? I don't know, I've not tested it, just speculating. But I would be happy to try it as it would be so convenient to be able to auto-connect/disconnect to BT-WiFi as I'm and about... Also, I am not sure if the App may be needed as a VPN client... (who knows *shrug*). BT need to be a lot more open about the opt-in to BT WiFi Community (and not just include it in the small print that realistically that nobody reads), I bet most BT broadband customers aren't even aware they are sharing their broadband (and they can connect to anyone else's). Thanks, Bruce.

Hi, Has anyone tried iPlayer on their now TV Box and notice that it's b0rked (since an update)? It's a shame because I bought it for iPlayer. Thanks, Bruce.

The Peak (e.g. Stanage Edge) for single or two pitches, Wales (Snowdonia) for multi-pitch....?

I did wonder about this and fortunately, I notice from the forums that BT do allow you to plug-in your own router, which prevents all these shenanigans with broadcasting additional SSIDs to share (although BT describe this as you will losing these features). Having said this, the Hub or whatever you call it seems very customization. You seem to be able to modify most things (including turning off the additional SSIDs). It looks like a lot of thought has gone into it. Now, if only I could turn off that blue LED... Thanks, Bruce.

It's interesting what you say about the laptop successfully establishing a 1000Mbit link and getting an IP lease (with no apparent problems I assume). I've seen it the other way around, in that 100Mbit devices connect fine, but 1000Mbit don't. The problem was that only 4 of the 8 wires were properly crimped in the socket. 100Mbit (100BaseT) requires only 4, but 1000Mbit (1000BaseT) requires all 8. One possibility is the printer has been briefly plugged into another socket (on another VLAN) and the DHCP server has leased an IP for it, but in another subnet. The DHCP server has spotted that a device with the same MAC already has a lease (but on another scope) and is refusing to hand out another IP? However, I still think it's more likely a wiring issue (layer 1) rather than a protocol/software issue. Thanks, Bruce.

I have just signed-up for BT Infinity (38Mbit unlimited), and so far I am happy with it. Getting downloads of about 33Mbit and uploads of 10Mbit. I have been automatically opted-in to the BT-WiFi network, where the customer (me) agrees to share a part of their bandwidth with other BT broadband customers (presumably who also opted-in). They connect to "BTWifi-with-FON" (being broadcast as an additional SSID) and authenticate on the captive portal using their BT username/password. I am happy enough to allow this as I can then borrow bandwidth from the 1,000s customers when I am out (without using up my mobile data which is currently 2GB). But I think BT need to be a lot more open about this. BT are also broadcasting a 3rd SSID from my Hub called "BTWifi-x", which is an 802.1x authenticated wireless network. After a little bit of research I have discovered that the purpose of this is to allow users to automatically connect/authenticate (rather than using the captive portal, which can be a real pain) to these wireless networks. The authentication relies on a digital certificate on the client device to authenticate the device as belonging to an eligible BT broadband customer. Great I thought, I would download the App/certificate to my Android and I can auto connect/authenticate to these wireless networks, without having to faff around with entering username/password. The only problem is that according to a forum post from BT, the App only works with the iPhone. There exists a BT WiFi App for Android but it doesn't allow the 802.1x auto connection/authentication. According to the posting, they looked at developing an App for Android found there was a lack of consistency in the way Android devices handle certificates and hence they have no plans to develop a version for Android. Has anyone heard about this? I have issues with BT's lack of openness about the BT WiFi community (sharing broadband, the auto opt-in, security, the purpose of the BTWifi-X network, who is eligible to use it). But also, if the issue with Android is trying to support older versions, then why not develop an App for newer versions of Android (e.g. 4.4 or above), as I assume the inconsistency stems from differences in older versions of Android. Are there any BT Infinity customers out there. Any thoughts? Bruce.

I just wondered if anyone has experienced any ARP issues within their layer2/3 switching (particularly involving 3Com,Cisco or Allied). Is there any sort of technology on 3Coms that could prevent/supress ARP requests or responses from reaching the destination or source? I am looking at perhaps broadcast suppressions or DHCP snooping (although the latter doesn't appear to be enabled). Could 60 DHCP clients (e.g. VoIP telephones) cause broadcast suppression to kick-in if rebooted all at the same time (DHCP requests)? This issue mainly occurs (intermittently) at the layer 3 switch's ARP table (entry missing for a particular device/PC). But I am not sure if the root cause is with the layer 3 switch or the edge switches. Any advice / ideas is appreciated. Regards, Bruce.

Just following on from the other sage suggestions/advice, you could also look at Microsoft Data Protection Manager. If the WAN links are fast enough, you could backup all you're data/servers at one site to another site over WAN (D2D off-site). DPM will backup MS-based workloads such as VMs (hosted by Hyper-V), SQL, Exchange, Shares. Backup is usually D2D (Disk to Disk). But D2D2T (Disk to Disk to Tape) or D2D2C (Disk to Disk to Cloud) can be configured. You just need one server and some storage (such local or as an iSCSI NAS). Licensing is cheap for Education. No native de-dup, but the incremental backups (think of off-site shadow copy) are very efficient and low bandwidth requirements. It's not difficult to setup yourself. The D2D2C is an interesting option as it natively supports Azure as the "cloud". All sorts of backup frequencies and policies can be applied to meet your DR requirements, such as backing up the a VM each night, but also backup the application/data within the VM periodically during the day. You could backup all servers to your NAS with a retention of say 14 or 28 days, then backup once every 2 weeks to the Cloud. Good luck with whatever you decide. Bruce.

I was just wondering if any people on this forum are using WOL on their network? If so, do you use off-the-shelf software or is it developed in-house? If you have PCs in multiple subnets, what method do you use to get the magic packets to all of the PCs? Do you use subnet-directed broadcasts from a single host or do you broadcast on each subnet at layer 2 (i.e. from a server/host on the same subnet), or even use unicast packets? What method do you use to determine the list of MAC addresses of the PCs on the subnet? E.g. Manually maintained lists, ARP enabled NICs, DHCP leases, or other method? We plan to use this to plan/schedule updates/installs overnight. And as a secondary objective, to wake up certain slow PCs 30 minutes before the users start in the morning (to speed up logon to desktop). Any advise/ideas are appreciated. Regards, Bruce.

...or not?

Hi, Just a quick question for a Friday morning, should the ports that Wireless Access Points are patched to be configured as Portfast? Thanks, Bruce.

I changed it a couple of weeks ago. I can't say I noticed much difference. Had no issues with the screenlock. One of the apps I use, Tv24.co.uk, doesn't always start properly now, which could be related. Also, scrolling in Chrome initially seemed a little unresponsive but now seems to be OK. Maybe a few glitches were fixed in 4.4.4, ready for Android L, where ART will be the default.