How To: Fix iPad loosing wireless network

How To: Fix iPad loosing wireless network

One of the issues being reported with the Apple iPad is that it looses it’s wireless network connection and then has a very hard time reconnecting.

I ran across this bug in the field where I had 6 iPads scattered around a football field, and two (or more) Apple Airport Extremes connected in WDS mode. The issue seems to popup usually when there are either a large number of people in between the Airport Base Station and the iPad, OR when the iPad was too far from one or the other Airports. Weirdly, the Wireless Connection icon in the upper left hand of the screen would have full bars, but still couldn’t get to any network devices or to the Internet.

From the iPad’s main screen, choose “Settings” then tap “Wi-Fi” locations. Choose the default Wireless Network, then tap the right arrow. Scroll down to “HTTP Proxy” and then tap on “Manual.” Back out of these screens and test the wireless out. If it still fails, go back and choose “Auto” instead of “Manual.” Note that there is no need to enter any information in the “Server,” “Port,” or “Authentication” fields. Unless your specific wireless network needs those.

This is reported to also work with iPods and possibly iPhones so the problem seems to be in IOS itself, and not with the hardware. Hopefully a future update will fix this issue.

Other Suggestions:
-Update Firmware on Wireless Access Point
-Turn off the 802.11n mode
-Temporarily disable security measures. If this seems to solve, set router to WPA2
-Disable 40mhz or 20mhz under 802.11n settings – especially in crowded wireless environments
-Set the iPad to get a static IP (only reported to work in one case, so very low on the list)
-Apple’s KB has a few other suggestions: http://support.apple.com/kb/TS1398

Home Network Setup

Home Network Setup

Originally Posted 11/20/2000. Ironically a lot of this is still useful.

 

Home Networking

 

 

            In the day and age of Multi-computer families, a frequent argument is, who gets to be online? The answer is simple, Modem Sharing or networking. Modem Sharing can be done one of several ways, using a regular Analog Modem (300 baud to 56.6K,) DSL or ADSL, ISDN and Cable modems. You could also be on a T-1 (or greater!) network connection, but then why are you reading this? J

           

 

Which one is right for me?

 

            Analog Modems

 

            Pros: Can be used anywhere a telephone is located.

            Cons: Limited to an upper bandwidth of 53K due to telephone regulations in the US. May be slightly higher in other countries. Modems also suffer greatly from phone line quality, crosstalk (where two or more wires touch each other in the wrong place therefore mixing the signal,) and any line noise at all. More then two low-bandwidth applications can easily make this almost worse then nothing. Ties up the phone line if you only have one.

 

            ISDN:

            Pros: Very fast, two 64K channels giving 128K total. Even one channel at 64K is cleaner and noticeably faster then 56k.

            Cons: Still only 128K, costs the same as DSL. In addition, totally against ISDN specs, some companies charge you for the 2nd line to automatically drop when an incoming or outgoing phone call is detected, so your phone is still tied up.

 

            ADSL/DSL:

 

            Pros: Very fast… minimum 265K maximum 7 megabits, makes this very, very fast. Phone lines do not get tied up at all. Can easily support multiple high-bandwidth applications (web browsers, games, large downloads, etc.) Can get a Static IP, which is very useful if you wish to run servers at home.

            Cons: Suffers slightly at the phone companies end due to misconfiguration. Such things as dropped carriers, being able to see other people’s computers (in Network Neighborhood.) and depending on the carrier, a slight slow down in bandwidth due to the phone company over selling it. Remember, at this point the phone company does not have to guarantee a certain amount of bandwidth… you simply have speeds up what ever your limit is.

 

            Cable:

 

            Pros: As above, very, very fast

            Cons: Suffers greatly from overselling of bandwidth. If other cable subscribers are all using it at the same time as you, they effect your speed. If you’re the only one in the area then it’s not a big deal. But frequently this slows down to speeds that are little better then a regular modem. Can’t get a Static IP number,  you’ll never be able to run a server, at least not very effectively.

 

Initial Setup

 

            First you need to get the network equipment. Your best bet for ease of configuration and portability is to get a couple of  mid-priced range network cards, (Kingston is a good bet) and a decent 100Base hub. (NetGear, Bay Networks, Intel.) You can buy cheaper cards such as SMC but they will effect your speed. If you can afford it, a switch is even better, giving you more bandwidth and speed. You’ll also want several lengths of networking cables too. J

            Second, your modem or router should be hooked up to one of the computers. In the case of some DSL routers, these can be directly hooked up to your hub. Other products have a DSL router and hub built in to the same unit. In my opinion this is unncessecary and could cause problems down the road. Remember to install the 2nd network card if it’s an external unit. (Cable and DSL only.)

            The third step is to setup the modem/router itself to connect to the network. This changes from different brands of Cable modems and routers. If you’re using a regular 56k modem, connect as you usually would.

            The last step is to make sure you know what your IP Address, DHCP Server (if needed,) DNS Server, Gateway and Netmask numbers are. You may or many not have all of these options. These have four numbers or “octets” like 255.255.255.0 or 192.168.1.0 Each of these numbers or octets can be anything between 0 to 255, with some exceptions.

            Your network should look something like this when you’re done:

            ISP -> Modem/Routers -> Computer -> Hub <-> Other computers

            One thing to point out, the computer plugged into the modem/router needs to be running an Operating system capable of doing routing. Windows 95 is not capable of doing this.  Windows 98 sorta does it, while Windows NT, ME, 2000, FreeBSD and Linux are all capable of doing so. Mac OS X being based off of FreeBSD can also do this. Windows 95 is capable of doing so with the help of a third party commercial piece of software. I personally run FreeBSD on  a 486 SX-25, recently upgraded to a Intel P-166 to do this. (BTW, it runs quicker and more efficiently then Windows NT or 2000 on a AMD –450 with 128 megs of RAM!)

 

Network Setup

 

            The Network hardware is pretty straightforward. One cable from each computer plugs into the hub or the switch. Each port should have a LED saying the port is active. Usually there is a 2nd LED that shows network traffic on that port of the hub.

            Start with the computer connected to the modem/router. The protocol that you wish to run is TCP/IP, make sure IPX/SPX and Netbui are both disabled unless you need them. Most games these days don’t care what protocol you’re running, but some of the older ones need IPX/SPX (Duke Nukem 3D, Warcraft, early Diablo releases, etc.) Having Netbui disabled helps keep people from hacking into your computer, there are hundreds of security holes in Windows with this protocol.

            After adding TCP/IP in the Network option under Control Panel Add in your network settings as provided by your ISP. Usually your ISP gives you instructions on how to do this, you’ll want to follow them. Test your dial-up and networking. The computer should run perfect.

            Now comes the tricky part, configuring the other computers. The easiest configuration is the dual network card and external modem/router option. The first NIC connected to the modem/router is configured exactly as it should be. The 2nd NIC’s TCP/IP Address should be one that is assigned by you. For ease you should pick something like 192.168.1.1 You can pick any number, but make sure the first three octets are all the same. Do not pick the same first three octets as the IP assigned to your 1st NIC… it will cause problems.

            Your Netmask is always 255.255.255.0 do not ever change this unless you know what you’re doing, your network will become unreachable. The Gateway address is the same as the IP address of the 1st network card. Keep in mind this is only for this card. Certain operating system will not accept this option either, but most Windows systems should.

            In addition some operating systems will need a flag set to make it a gateway or router machine. For instance, in FreeBSD and Mac OS X /etc/rc.conf needs the line “gateway=enable” This also invokes the NAT or “Name Address Translation” program. NT and 2000 should both give you a similar option.

            The 2nd (and all subsequent computers) will have an IP address of 192.168.1.x (where x is any number between 2 and 254, do not use 1 as it’s already used, 0 and 255 should also be reserved, Just in case.) Their Gateway address will be the IP of the 2nd network card in the first computer! DNS or Name Server address will be the same across all computers, unless you setup a nameserver, which for a small network I recommend against doing.

            For computer with a single network card, and a modem, the theory is the same. Use the gateway to be the IP number of the modem/router. Remember that IP numbers are assigned to Network interfaces, not to the computer. A computer can theoretically have an infinite amount of IP addresses, in reality it’s whatever number the operating system is capable of handling, at least 254 though, and more then most people will ever need.

            If you have a large amount of computers, say more then 5 or 6, you may want to look into assigning these IP numbers via DHCP, it’s initial setup is harder, but more robust with large networks, it also makes the client machines virtually Plug and Play.

 

            Security

           

            Now that your network is running, it’s time to turn your eye towards security. The best way is a firewall. Windows NT, 2000, FreeBSD and Linux can all do this. Linux has more security holes fixed then Windows NT or 2000, and FreeBSD even more. If you’re really serious about security and wish to really learn, OpenBSD is the best of all. FreeBSD offers the best choice of security and ease of setup.

            A Firewall is simply that, it keeps the bad stuff from coming in, and if you wish, keeps stuff from going out. It works by stopping or dropping traffic if it does not adhere to the rules that you define.

            The first step in building effective firewalls is to block everything. Then step by step you want to unblock stuff you want. Port 80 (http requests,) 21 and 22 (FTP) are good starts. Some programs will want other ports open, PCAnywhere needs 186 & 187 open. Remember, it’s easier to block stuff before it’s broken, then to block if after somebody has hacked into your computer.

 

            Conclusion

 

            These are really some very basic instructions. Unluckily due to the sheer amount of different configurations out there, I can’t give step by step instructions like I’d like too. If you have specific questions though, post to the Hardware and Software forums and somebody can help! J Even then you should have a working network configuration in no time at all!

How To: Budget Network Attached Raid 5 Fileserver, Part 3

The server is up and running, the drives are formatted and now serving just under 750gb of Raid5 goodness. Since I have a mixed network of machines, Windows and Mac I need to make it will work with each of those. So we’re going to setup SMB. I could also setup NFS as it’s slightly faster and would add compatibility to Linux machines but it’s not needed in this case. And SMB is actually a more common need since most people run Windows.

Again, counter intuitively I found it was best to setup users first before trying to setup the actual shares. There are two ways to do this. The easy way, or the secure way. If the NAS server is on a local home network, guest access may be the way to go. But I always suggest even in such cases to at least have basic security.Go to Access then Users and Groups. Type in the name of the first user and a password. Leave the User ID to default, it will increment automatically for every new user created. The next option, Primary Group has a drop down list of the standard Unix groups.      

Users and Groups   

The first user should be added to the admin group. Other groups can be staff and wheel. This will give the user maximum permissions to read and write files. Again for home use this is sufficient. Of course FreeNAS also has access to add LDAP and Active Directory authentication. Both of these are simple setups for any System Administrator.Since it’s first on the list, and there are more Windows users around, we’ll explore setting up SMB before NFS. Under Services go to CIFS/SMB. Click on the Shares Tab, then the Plus to add a new share. I simply called the name raid and put in a comment to note which raid it was. (In this case, raid because it’s the only one.) On the path option click on button with three dots. This brings up a filebrowser window, chose the raid directory that we’ve already setup and hit ok.  

File Browser  

 As can be seen below the it automatically puts in /mnt/raid which is what we setup back in Part 2.  

shares   

Before clicking the OK button notice two options for increased security. Host Deny and Host Allow. Adding in specific IP addresses to the allow box and ALL for deny will greatly increase security for connections. Keep in mind if the IP address of the computer changes, for instance DHCP on the DSL/Cable router is reset this could cause connection problems later on. At this point, the server is ready to test. Mac OS X 10.5(Leopard) actually sees these connections automatically, even if sharing across SMB (under file sharing) is turned off. Mac OS X 10.4 (Tiger) will need to connect to the drive via the Connect to Server option. Windows can now be connected just like any other server. Vista Directions are here, and Windows XP is here

FreeNAS is an incredibly powerful program. Even though it’s still only in Beta it seems like a very solid product and does RAID NAS very well. I’ve really enjoyed playing with it and want to give a big congratulations to the development team for a job well done.

How To: Budget Network Attached Raid 5 Fileserver, Part 2

How To: Budget Network Attached Raid 5 Fileserver, Part 2

In Part One I talked about selecting the hardware and my thoughts behind the choices I made. It’s all been put together and the next step is to install an OS. For various reasons I decided to install FreeNAS which is based off of FreeBSD. It can be grabbed from here.

After a few minutes of looking at FreeNAS I was really quite impressed with it. They’ve taken a quite hard and convoluted process, added menus and made quite easy to setup. Like FreeBSD in general it can feel picky about hardware. If you’re using some off the shelf no-name SATA RAID controller, the odds are it’s not going be supported. But a lot of the more popular and better quality models are supported. The FreeNAS website (and also the FreeBSD website,) are both a tiny bit hard to find information and support when first using it. Counter intuitively you need to click on the Wiki link first, then knowledge base, not the Support link to find Installation and Configuration documentation. Luckily the menus within FreeNAS are fairly self explanatory.

The first step is of course to download the image. In my case I grabbed the live CD so that I could simple have the machine boot off of it and was good to go. Another option is to use a USB thumb drive to boot off of. I’m personally disinclined to use one as they stick out and get broken easily.A UNIX installer screen will come up and start probing and self configuring hardware in the machine. A FreeNAS graphic screen may come up, and eventually it’ll beep when ready. Hit the escape key and choose option 2 to get an IP address via DHCP. Make sure to hit “Yes” when it wants to choose a IPv6 address. That step messed me up the first time I saw it, but it’ll simply fail as most likely there is not an IPv6 server around. Most home routers have a DHCP server built in, but there may be some configuration needed so check the router’s documentation.

Once the IP address has been discovered, type the address it gives you into a web browser to open up the FreeNAS configuration page. The default user name and password is admin:freenas, it’s highly suggest you change the password ASAP. Once everything is fully configured we’ll go ahead and change this. At this point the instructions proved useful.

Step One is to add the physical disks. Under Disks Click on Management, then the + sign. This brings up the disk management screen:

As can be seen the available disks are at the top. In this case ad0 is the 40gb IDE drive I’m going to eventually use as a boot disk. Per the instructions change the “Preformatted file system” option to “Software raid” (the other options in that article may not be available). In this case I have four SATA drives so each needs to be added individually. Hit the apply button and each drive is added.

The next step is to create the raid partition. Go to Disks, Software RAID, then choose RAID5. Choose a RAID name it doesn’t matter what. For simplicity I used “server” put a check next to all the disks that are going to be part of the RAID then another on the “Format and Initialize” box. Hit OK, then once again “Apply”. Now sit back and wait. On this screen very little is going on. But on the RAID server itself messages will start popping up. Even better, they’re helpful!GEOM_RAD5: server: all(-1): re-sync in progress: 0.01% p:x ETA:232min (cause: store verify progess). After 232 minutes of waiting we then got this screen:

As can be seen my four 250gb SATA drives have been built into a 715gb RAID5 Partition. It now needs to be formatted which is done in Disk, Format. Choose the RAID array, give it a name again, no hurt in using the same name again. We’ll format it out to UFS+ with GPT and Softupdates as the as the filesystem. The other options may work, but are not recommended by the FreeNAS team. Hit the format button and thirty seconds later the drive is ready to mount.

The mounting screen is sort of confusing at this point. After all we’ve already created and formatted the RAID drive so it should be ready. But this physically mounts it so that it can actually be accessed.

Under Disks, click on Mount Point. Pick the disk. Then choose “EFI GPT” under the partition menu. This menu was a bit confusing for me at first, and once again FreeNAS’s documentation left this step out. Reading it at first it seems option 1 was wanted as we’d setup UFS before. The filesystem stays UFS though, and the name can be what ever. I choose the simple “raid” moniker for simplicity. The last option could be a real life saver if the power ever goes out. “Enable foreground/background file system consistency check during boot process,” would run fsck and other filesystem utilities when the machine was powered back on. It might take longer to get the RAID back up, but could save problems in the long run.At this point we’re ready to start mounting the RAID and writing data to it. I’ll talk about doing that in the next article in this series, including troubleshooting and setting up Time Machine to back up to the RAID automatically.

Part Three