Monday, July 18, 2016

FortiGate Huawei E8372



There doesn't seem to be a ton of info around 4G/LTE USB dongles in FortiGate devices at the moment/ever, and the FortiOS GUI has never really been fantastic for this, thought I'd start dumping my own experience just in case it may help someone else again.

Before the  Huawei E8372 I'd only ever tried the Sierra/Netgear AC320U along with a handful of low end consumer 3G usb sticks from Huawei. All had pretty mixed results and required a ton of searching to find the correct config and specific initialisation strings to get everything running smooth. The E8372 seem to be a lot easier.

I've tried this on FortiOS 5.2.8 firmware and didn't have any luck. I didn't try and find out 'why' so your mileage may vary. I'm also no forti expect so help me out if you think something could be done better.

Firmware 5.4.1 (I performed all my tests on a FortiWifi-60D)
Huawei E8372h-608 software version 21.180.19.06.203 (Telstra)

Jump into the CLI and enter the following

config system lte-modem
    set status enable
end

Because the E8372 has everything configured within itself (its a full router/wifi ap/modem) you shouldn't need any APN or init strings (when I tried an APN string it just ignored it and continued to use its internal settings), the Forti will then leverage (what appears to be) the NDIS type/virtual LAN connection of the E8372.

You then see it appear under interfaces like this (I had already place mine in a load balance config before this screen grab)


It terms of speed I can easily flood the USB 2.0 port of the Forti using the E8372, easily getting 40+Mbps, and it seems to be stable for long periods of time, but I was having some issues when the Forti powered on.

After powering on no traffic would head over the link, and while the interface appeared up no DHCP lease from the E8372 would be received, or if statically assigned I couldn't ping the address of the E8372.

Dropping (shutting) the interface and bringing it up again after a small amount of time (this was important) seemed to fix the link in most situations.

config system interface
    edit usb-wan
        set status down
        next
    edit usb-wan
       set status up
       end

Hard to say how much time it needs to be down for, but I found if I didn't have it down for at least 20 seconds the method wouldn't work.

I have no idea why the Forti would be having this issue, maybe there's an init string that will help? For now this seems to be a solid fix.

I chucked the above into an auto-script to run every time the Forti turned on.
   


Thursday, June 07, 2012

Venus Transit

I managed to duck out of the office for a few seconds yesterday with Dev Ben to have a look at the transit. We just made it at 1:50pm, and could clearly see Venus in the bottom left corner. Amazing to have that kind of view of another planet. Ended up viewing through a welding mask and a pair of dark safety glasses underneath (welding mask long was good, but you just needed that little bit more). Very cool. Glad I had the chance.
 





Tuesday, May 22, 2012

Raspbmc (xbmc for Raspberry Pi)

Nothing crazy to report, just having fun with it.

Getting really nice playback performance of 1080  mkv files. Looks very promising.

Friday, May 18, 2012

Raspberry Pi

Well this is a little exciting, I wasn't expecting to get my hands on a Raspberry Pi for another month or so. Apparently it's one of 54 in Australia so I'm feeling pretty lucky at the moment.

Time smash some quake bots...

Tuesday, September 27, 2011

Easily Secure Erase an SSD (alternate to HDDerase.exe)

EDIT: OCZ have their own secure erasing and firmware updating Linux distro now, and have a great 'how too' on their forums. Check it out here

I was recently looking for a way to run a Secure Erase command on my SSD to ensure its performance wasn't being hampered. I tried a ton of different ways (including HDDerase.exe and hdparm) but all failed due to the security on my mother board (and laptop) that stops certain harmful commands like Secure Erase being issued to a hard drive.

If you don't know what performance issues I'm referring too then you should stop and do a little reading. AnandTech has a great article on it here, and just to be clear this post has nothing to do with data security on SSD's, its only about performance.

If you're like me and have had trouble finding an easy solution to issuing a secure erase command to your SSD's controller then this might be a solution for you.

You will need Parted Magic for this (free and opensource), and obviously ability to boot to it. I used 6.6 from a CDRom but I'm positive it will work just as well from any other version and any other medium (USB, network).

Warning, this will completely erase your drive, be careful! If you have important data on other drives in the same system then it you should probably unplug them, just in case :)

1. When booted, hit the menu button in the bottom left, head to System Tools and then Erase Disk 

2. A menu will be displayed like the picture below. You want to choose Internal:Secure Erase command writes zeroes to entire data area (Parted Magic is using hdparm to perform this)

3. Choose your disk


4. Here is the tricky that Parted Magic handles much better than the rest of the utilities out there. You need to tell the computer to Sleep so that, when it wakes, the drives security will be unfrozen.

5. Wake the computer up, go through 1, 2 and 3 and you should have a different menu instead of what was displayed in 4. If you don't, then run through the process again. If that doesn't work then your system may be e a little more secure then most and this probably wont work. Don't change the password, just leave it as NULL (unless you know what you are doing of course)

6. Confirm the erase

7. And your done, and yes, it is normal for it to only take a couple of seconds. On a normal platter hard drive it would take longer, but this is normal for an SSD. Its faster because basically SECURE ERASE command just resets the SSD's controller instead of erasing each sector, like it would on a platter based drive. If it take longer than 30 seconds you have an problem or you have chosen a platter drive.

8. Jump into GParted to confirm your that the erase has worked, your should have an entirely blank SSD. If it still has a partition then the Secure Erase command hasn't worked or you have erase the wrong drive. 

This is a really easy, GUI way to use hdparm utility to issue Secure Erase command, and the 'sleep' trick seems to work on everything I've tried (ASUS P5KC, MacBook Pro 6.2). 

Sorry for the bad photos, I did this in a rush. 

Friday, August 26, 2011

Is HP really killing the Touchpad, or is it all a massive ruse?

Back story here and here


I managed to get my hands on two HP Touchpads earlier this week, both at the crazy prices of $98AU for a 16GB and $148AU for a 32GB (finally, Australia has cheaper tech than the US). It was a mad rush. If a work mate and I weren't watching twitter for rumors we would have been none the wiser. 

We got to the local Harvey Norman store around 15 minutes after their staff got the call from the head office instructing them to go ahead with the reduced pricing. Five others were already waiting for the stock to be brought out and more steadily came walking (fast) in behind us. They had 70 in store and they were all gone within the hour.

Was it worth it? Completely. WebOS does lack a lot of polish and some really important features, but it's really not as bad as what some are making it out to be, and it's certainly not bad enough to kill the entire product line so soon. It had only spent four days on the shop shelves in Australia before the big axing. 


You can easily see that a lot of hard work has gone into the OS over its short life. Some of the WebOS aspects actually thrash the rivals, like multitasking. It's absolutely a ton better at multitasking than its rivals. I think that if it was given the same time that Android and IOS have had to mature I think WebOS would have been (and possible still will be) a very strong player in the ARM operating system game. 

The hardware is great. Again its missing a few features but its one of the fastest tablets (CPU and GPU) in the current market. It feels strong, looks strong, was packaged well, has a good screen and was released along side some high quality accessories. I quickly grabbed covers for both and a touchstone charger in fear they would be hard to get after a few days, which another work mate is unfortunately confirming right now. 

It all seems crazy to the point where I'm starting to question HP's motive. Six days after they had internationally killed the entire line they released/updated a HP branded application which is specifically designed for the Touchpad. Then today they have switch on paid app support for Australia. Not bad for a "completely unsupported software" platform. Why would they bother? To make it even more confusing HP have just rewarded Touchpad owners with a free "6-Pack" of paid apps in the U.S. 

If HP are trying to show possible buyers of their consumer line that they still have confidence in the Touchpad/WebOS product its a very extremely weak attempt, especially after their drastic and sudden actions to wipe their hands of it all. 

Whats with this half/lingering interest HP? I know I'm not the only one who is thinking along these lines. Could HP turn around now and say "now we have hundreds of thousands of users on the platform we will continue with the line"? Sony and Microsoft sell their consoles at a loss at the start of their lives to get customers on the platform, is it that hard to imagine a company like HP could be doing the same? If I was a hardware manufacturer and I'd just bought a reasonably fresh OS I would certainly think about doing it. Maybe I'm just being over inquisitive/over speculative...?

Either way I think I have won. I'd put $100 on an Android build being publicly released before the end of the year. If that doesn't happen I'm happy to continue to use WebOS in its current state. 


Monday, July 18, 2011

BIOS Update for Sunix SATA4000 (Silicon Image 3114) PCI Raid Card

I've had one of these Sunix SATA4000 PCI Raid (fake raid) cards for a while now, and although they are pretty limited in terms of abilities and performance they are reasonably reliable (500+ days on mine without a problem) and very cheap (I paid $10 delivered on Ebay). They use a Silicon Image Sil3114 chip and BIOS so software is pretty easy to get for it as well link.

It's one major drawback is the card ships with a much older BIOS than what's available through Silicon Image, which cant recognise larger hard drives (1TB and up). You can upgrade the BIOS using the Silicon Image BIOS and tool but you need to know what flash chip the cards uses.

The original post here that I found (which has been removed for some time now) stated that it used an AMD 29LV010B 1MB compatible. Using that worked fine for me and now I'm running a 2TB drive on the card without any problem. I also noticed around 20-30MBps more in IO to the older drives on the card. I thought I'd make a quick post about this just in case someone else wanted to know, references to it seem to be disappearing quickly.