KK20

written in house (by me too). Our SMS system, after school club scheduler, school bus timetable system (that the bus company updates for issues), exam seating planner (we have people who are not on our MIS but external local candidates etc), staff holiday planner, pastoral tracker for our commendations, merits, credits, house colours etc those sorts of things (they all integrate into moodle). As for our MIS it is primarily used by our admin and SMT. It would be unlikely we would be using a large amount of MIS "parent portal" type screens as we already do what we need in moodle. The above MIS were chosen due to the modular(ish) nature so we wont need to buy/pay for the bits we wont use. I only use "read only" data for my apps and am happy to use an intermediatory table (as I have to in E1. E1 is horrific to get data out of - but it is possible!) Just wondered if anyone had used the above and the pitfalls/strengths of them.

we are independant. The database must be accessible as we have other programs that utilise the underlying data.

I have never used ruckus but I assume they are VLAN aware APs (I would expect so since cheaper unifi ones are). As others have said, it doesnt matter what subnet your existing vlans are on your switches, your "new" vlan will be treat separately (the whole idea of a vlan). So as long as you have (say) vlan 50 set up on every switch that needs a ruckus AP and you have with tagged uplinks and the ruckus switch port as tagged to vlan 50 (with default vlan left alone and probably still set to untagged on whatever your default PVID is). I imagine inside the ruckus management software you have your "normal" SSID set up and can add "guest SSID" but locked to VLAN 50. Have your L3 switch route as appropriate and the voodoo will work. If you want to test the setup easily, set up two PCs with static ip addresses different to any of your network subnets (we dont use 192.168.x.x so that is ideal for my tests). Set up a port on one switch as untagged VLAN50 with PVID of vlan50 (it wont be a member of your "default" vlan), do the same on another switch elsewhere and they should be able to talk to each other without leaking out onto your network. If that works then your guest wireless should work too (dont forget to either not use your test ports or put them back! It is so easy to forget this and wonder why PCxxx isnt connecting to the domain...!)

Inherited a network. Core is Netgear 748's, 724's and 2x 748TS in a cabinet network is about 2/3 star. a half dozen satellite 748 and 724 elsewhere that are on either 1gb fibre (OM1) back to core or single 1gb copper back to core. No SAN. Each netgear (apart from the 748TS which were interlinked HDMI) were daisychained via a single copper cable each when I arrived (!). The servers werent even on the TS either(!!). First thing I did was aggregate the switches with as many LAGs as I could spare ports and rearrange the heavy load (i.e. ICT rooms etc). They were using TMG to route between "default" VLAN (no management VLAN just default for everything) and a couple of other VLANS used for a guest wifi access point and the finance machines. No L3 router. An access point on each VLAN was used as a dhcp server. After setting up DHCP on server 2008 to act as DHCP I am now going ot invest in an L3 switch and start working towards getting a SAN and virtualise (only 2 servers + one small firewall TMG server)/cluster for redundancy over the summer. Ive not had time to sort out the VLANs yet and get everyone off default since everything is netgear already i'm quite happy to get another netgear L3 switch but what I have no experience on is linking netgear kit between their L2 and L3 offerings. Will the 748TS HDMI link to any other types of switches? Ideally if I can do that it will save me burning up precious ports on the L3 by LAGing them - I am looking to the future and whilst I might not be able to afford 10gb links to the SAN/server initially (LAG will have to do this year if the budget wont stretch) I might as well look at using 10Gb as soon as I can. Any recommendations on netgear kit? Is it more cost effective to get a "normal" 1gb L3 switch and an L2 core server storage 10Gb capable switch?

Inherited E1. I am convinced that E1 is endorsed by satan. Have the go-ahead to look for a replacement. Ive narrowed the choices to Engage, iSAMS and Schoolbase. Anyone use the above? Anyone loathe the above? Major prerequisite is that it is hosted by us and we have access to read the database independently (SQL would be ideal).

we use redSMS via their API from a little app based on our MIS database. We tend to buy a block of 10,000 SMS per year which is good enough for us (about £480) 500 students

our IPSCA did the same - failed on email but passed after human intervention. Not had an issue with modern browsers although older mobiles ones do fail. We added the cert to our early ipads later ones are ok.

Sort of shows the obvious. You get what you pay for. Pay ruckus/aerohive prices and get the best. Pay unifi prices and dont expect to keep a class of 30 GDIs running smoothly off one AP. What is interesting is that I do run 10 ipads streaming off one unifi AP (probably not the same video type etc) but it isnt a pro and they are set to HT40 radio. According to this report the unifis got a proper shoeing.

I thumbed through it, seems that v3 has better guest portal, better load balancing, better hand off, uses existing hardware. No plans for AP updates over the air (so no true AP bridging or repeaters). New hardware on the horizon but nothing to disclose. No plans for > 4 SSIDs although it is technically possible.

I'm a recent convert to unifis. I have 15 now (5 packs of LR) all performing brilliantly. I dont have 100% coverage yet but I cannot sing their praises enough. We dont have more than about 15 devices connecting to each AP at any one time and have very little overlap on coverage so I cannot comment on their load balancing. Certainly roaming and handoff appears to work on laptops (not so on android phones but that is a problem with the stack they use apparantly). Guest isolated VLAN works - once you figure out they need access to the controller on the VLAN subnet so routing of the controller ports was necessary (we didnt have an L3 switch - only smart L2 core switches- in the early stages so that was "fun"). Now we have an open guest network with the guests using routed DNS to our internal server, intranet redirects so it doesnt go out on the internet and VPN access so that staff can use their own laptops with VPN to our network if they want. Vouchers are used via the hotspot manager for "1 hour" duration (heads of department can hand these out to students for specific purposes), 1 day for visitors, 1 year for staff and a couple of "permanent" ones for exceptions (my phone and laptop etc). I have an isolated VLAN for an old fashioned WEP SSID used for a couple of legacy devices, that VLAN connects only to the server required so insecure WEP is isolated from the rest of the network. Then there is our "normal" WPA2 SSID for our domain machines. I suppose I could theoretically run out of SSIDs now! DHCP is relayed to our DHCP server for the VLANs and routes added for the unifi controller ports Overall they are a fantastic bit of kit for the price - although they do need a bit of work (and some forum voodoo searching) to get working fully.

just look for a printer with a parallel port and assign the printer to LPT1 with a generic LPT driver. There are loads of second hand refurbished printers that would suit.

bump for great justice! (yes I know it is a 3 year old thread but damn useful!) Got a pupils USB "not formatted" working again, knew about recuva but not pc inspector. Obviously they had never used their my documents (which is backed up of course).

for our exchange 2010 we use or internal and external name plus autodiscover. We choose to use a different OWA address. if you dont care about autodiscover then you might be able to get away with mail.DOMAIN and use DNS workarounds internally. We use "certificates for exchange" SAN cert.

3/1 from my mobile. Internet here is a little more pedestrian at 10/1 second picture is the fastest I've had from my mobile. Both on 3 network. Connection tends to hover around 3mb

Aha cars and bikes. Pretty much where all my money has gone over the years. Passed my test back in 1990. Skoda Estelle 1200 Landrover series 2a Fiesta 900 + CBR600 Astra 1.8 (was nicked within 2 weeks) + CBR600(still) BMW 325 + CBR600 still Laguna 3.0 Mondeo ST220 + CB600 (on account of my back not being as good as it was) SMAX 2.5T + CB600

"Thank you for contacting Barracuda Networks. I have looked at domain cumbria.sch.uk and I have requested for it to be re-classified. Please allow 2 to 12 hours for the change to propagate through our system. Thank you for your patience," Should be sorted soon (hopefully)

Only found this by accident through one of our suppliers. It seems that any school in cumbria using cumbria.sch.uk will be blocked by barracuda. Grrr. We have our own email servers and are squeaky clean so although .cumbria.sch.uk is free from any blocklist barracuda will block it as a subdomain. Hopefully their [email protected] is monitored and actually has someone reading false positives....

Fujitsus need an ACPI driver too, nothing critical though.

I stumbled on this post by accident. I changed next to no GPOs when I started rolling out W7 clients. Folder redirection simply worked for us. Libraries worked and some extra folders came up with the cannot open due to restrictions blah blah which was perfect. Student could save happily in their "my documents" (libraries/documents) etc. We use mandatory profiles though and I think this is possibly the key.

We use a WPAD solution to sort our IE and WinHTTP client/proxy configuration. It was JRE that caused the issue - I remember now. JRE *should* use IE settings (configured to do so and does by default) and can be seen to be using the correct proxy (via DG logs) but still cannot get the CRL resolution sorted internally within JRE. Some apps just wont play nice.

I had something similar to this a while ago (Java updates spring to mind but I cannot remember). I traced it back to windows checking revocation lists but not using the proxy (i.e. trying to go straight out rather than use I.E. settings like a good boy). Since some applications we have just dont play nice with proxy settings I ended up installing a proxy redirector on our gateway that went to a non-authenticated "whitelist only" squid proxy. I needed to add the CRL address onto that whitelist and all was good.

Sorry I forgot to say, I *did* set the host OS to windows 7 before upload so probably the same as you (i.e. have no issues). I only use a single partition on my PCs so I simply tell unattend.xml to expand.

As an aside, I have created my W7 image within a VM. I have had no issues imaging and deploying this through FOG. Initially I forgot about the 100mb system partition and set fog to upload the image as SINGLE partition. this still worked just fine on my first test PC. I since then use multiple partition model within fog and all is still well.

Whilst this might be a grave dig I was lead here from a google search. Mods feel free to split as appropriate. I needed to look into ways of automatically upgrading existing systems to windows 7. We didnt have any spare money so extra licences for servers or SCCM was out. That pretty much left us with either clonezilla live or FOG. Previoulsy I have used clonezilla live server to capture sysprepped images and deploy semi manually to systems (then domain join manually. For this project we have FOG server running under a virtual machine (ubuntu 12.04). The images are linked to our NAS storage drive (as an NFS share). Once up and running FOG is pretty easy to use - I would recommed avoiding quick registration though unless you like wading through mac addresses. Our windows 7 image was created using this tutorial Windows 7 Deployment FOG- SAD2 Driver tool > FOG Forums > Open Source Computer Cloning Solution and is a "universal" image. Has worked on our dell 330, 380 machines. Has also worked on our fujitsu A530's and toshiba 660 fleet of laptops. Obviously it relies on network pxe boot (so password that pxe fog screen options!) and also needs to do a little configuring with a C module (for AD security). All this is well documented and works but FOG is not a turnkey solution! I do not use any other FOG services so I cannot comment on those. I had made a minor mistake with the image (missed a piece of software off which I could have deployed via GPO if necessary) so it was a great way to test the refresh image (and rejoin AD) portion of FOG for a room full of PCs. It worked and I was very (pleasantly) surprised!

IE8 doesnt cache it. IE6 used to *occasionally* I cant say I ever tested it with IE7. proxy.pac is the way to go and the only way I sorted a few issues out internally here.