Browsed by
Category: Recent Projects

Customer Configuration of Calix 844 GigaCenter’s

Customer Configuration of Calix 844 GigaCenter’s

We have been deploying Calix 844’s for the past few months at SandyNet, and we have had almost no issues, and have received almost no complaint regarding the units. Calix did a great job designing these units, and in my personal opinion, they are a large step up from the 836g units. Receiving them almost hot off the production line, we quickly rushed these things to deployment, and I will admit, I think they are one of the big reasons our fiber deployment has been so successful so far.  From a customer standpoint, they need basic services such as reliable fast internet, dual band WiFi and an easy to use UI. The GigaCenter combines all of those main features, and more into one slick looking box. We can even remote manage these boxes so if a customer does not know how to change their SSID, or security key, we can now do it from the comfort of our office(through consumer connect), instead of making a house call for a five minute fix. So far, we have been extremely satisfied with these ONT’s, since they simply work. Time and time again, at SandyNet, we have acquired different devices in hope of finding a solution that simply works. We did not find it in Ubiquiti, Ruckus, Proxim, Mikrotik, etc. Mixing and matching these pieces of hardware in a production environment can sometimes result in a lot of problems (Like trunking between Mikrotik and Cisco). Building a stand alone Calix system from the ground up, has been an overwhelmingly pleasant experience. Okay, I will admit, it was rough until we understood the system and how to configure it, but once we got our bearings, it was pretty smooth sailing. Some minor issues were firmware bugs that were quickly resolved. And we mean quickly. We’re not talking about firmware upgrades that take six months and are more detrimental than useful, *cough* UBNT. Simple problems with memory leaks or incompatibility with 844’s and some new Macintosh’s were resolved in a very reasonable time.  Firmware upgrades are a breeze, and overall, having such a great ONT has already saved up so much time, and has allowed us to focus on other projects and problems such as continuing the deployment of our fiber network.

From a technician standpoint at SandyNet, the provisioning of an ONT is very easy. We pull the unit off the shelf in the morning, and input the FSAN into CMS and record it in our database, and send the unit out with our installer to be placed in a customers home. Once installed, the ONT upgrades its firmware, reboots and then applies its configuration. That is all there is to it. The customer is now online. One unit contains Ethernet, WiFi, RJ-11 ports and modem into one box. Customers no longer need to purchase a wireless router, and most of our deployments contain only an 844. The simplicity of the device makes it mighty attractive to our customers and us, since it is less complicated for the customer, and it is all squeezed into one device that we can manage easily. Most of our customers do not know how to change their WiFi options, or what the best practices are, so they often call in after being installed, requesting us to help configure the device. Other customers have enough background knowledge to login locally and change their options. Either way, the setup mo-betta than our previous infrastructure.

Now, all I have done is praise Calix for these units. That is not all I intend to do. Some customers have searched for help in configuring their ONT 844’s, and that is what is next.

Configuration of the 844-

Our installers should be placing a sticker on the ONT that lists the factory default settings for the modem. It should include the SSID and Key to connect to the WiFi, and the default IP for the web interface on the device. Below that, there is a username/password that is the default login info for the device. Before customizing your modem, complete the following.

Connect the computer you are using to configure the modem to one of the Ethernet ports on the ONT. If you are changing WiFi settings over WiFi, you’re gonna have a bad time. Once you are connected through a patch cable, you may open up your favorite browser and navigate to the gateway IP (Generally 192.168.1.1) and input the username and password (Username: admin, Password: *checkthesticker*). You should be greeted with a friendly looking page like below.

calix-homepage

 

From here, we have a few options:

Status – will show you information regarding the unit, and its many status including devices, configurations of WiFi and any associated devices.

Quick Start – is a simple configuration wizard that helps customers quickly configure their ONT.

Wireless – provides all options for configuring any WiFi related feature.

Utilities – provides troubleshooting programs to help determine possible problems, or view log information

Advanced – holds all of the less common options for ONTs including port forwarding, QoS, Routing and Network options

Support – provides details when receiving help from a SandyNet technician

Most of the configuration will simply be done under the Wireless tab, since everything else is pre-configured, or not commonly changed.

calix-wireless

Under the Wireless tab, there are four side menu buttons, 2.4G Network, 5G Network, Advanced Radio Setup and WPS. For the sake of making this simple, we are only going to operate within the 2.4 and 5 G Network buttons.

Note: 2.4G is currently the most common frequency for WiFi, so this radio should probably be used. 5G is standard on all devices within the past couple of years, and can be enabled if your devices support it.

First, lets make sure we have the 2.4 radio turned on. It is on by default, but select the Radio Setup button under the 2.4G Network button the left hand side. make sure the Wireless radio is set to on and now off. Hit apply after you have made your change.

calix-wireless-radio

Next, lets give our 2.4GHz network a good name. Select the SSID Setup tab and select the SSID that is named CXNKXXXXXXXX and make sure it is enabled. Now you get to be creative (or not so creative) and change the name of your wireless network. A lot of our customers want to keep their previous WiFi configuration, so if you wish to do so, fill out the Rename SSID box with your previous wireless network name (It is case sensitive). If not, come up with a good identifier for your WiFi, and no, FBI-surveillance-van-3 is not a good name, since everyone seems to set their WiFi to that.

calix-wifi-settings

 

Hit apply and lets move onto security.

Under the security button, you will need to now select your newly renamed SSID from the dropdown menu labeled SSID (Network Name). Now we get to select the security type.

The following options are available:

WPA-WPA2-Personal – Combines both encryption methods of WPA and WPA2 for maximum compatibility of devices. This will except the passphrase in both encryption levels, making is less secure than WPA2, but most function for all devices.

WPA2-Personal – The strongest encryption method for WiFi at the time of this article. Any non-compatible WPA2 devices will be unable to connect, so make sure all of your devices are compatible.

WEP – Is extremely weak, and in my book is not an encryption. There is no algorithm, just a HEX code encrypting the data, making it extremely unsecured.

Security-Off – makes the network open for anyone to connect. There is no password.

Pick your desired security type and then move on to the encryption type. For WPA and WPA2, I recommend AES, since it is the best. For compatibility you may enable TKIP or both.

Now you can set your security key. If you wish to keep the ridiculously long default key, be my guest, but most people want to name it to their pet or something easily guessable. Hit the button Use Custom Security Key and type in your key. Hit Apply when you are done.

calix-wifi-security

 

Woot! you have configured your 2.4G network to be whatever you wanted! If you wish to enable the 5G network, do the same thing under the 5G button.

 

As you can see, the Calix GigaCenter UI is very easy to use, and pleasing to the eye. Configuration of WiFi is extremely simple, not to mention its super dooper range! 🙂 We are happy with these devices, and we believe customers are too. Like always if you have questions regarding me, my poor humor or how I became such an awesome person, email me at gbrewster@agoasite.com. But if you are a customer in need of help with WiFi, SandyNet or the City of Sandy contact them, not me at: support@sandynet.org or call 503-668-2923, and you might get me on the line!

Getac 4G Card Woes

Getac 4G Card Woes

We have had our Getac F110’s in production for about five months, and while we have not had issues with speed of the computer, we have had numerous reported issues of cell signal dropping and not being able to connect unless the tablet was rebooted. Often saying no device detected, or searching… Now, we are using Windows7, and when I first imaged them, they featured the Sierra AirCard watcher utility.

skylight

In September of 2014, Getac released SkyLight, which replaced the AirCard watcher, and did exactly the same thing. It even looked the same. This however, did not resolve our issue. On top of that, it made it harder for us to deploy new Getac’s, since it didn’t seem to activate newly install SIM cards. Well I am here to say that after a lot of frustration, I have found out why, and made it really easy to remedy.

On October 29th, 2014, Getac released a Jpeg called 4G module driver with the description – How to check 4G module Firmware. The image has been uploaded below.

WWAN check FW (1)

 

Note: When we imaged our machines we took all of the drivers from the Windows 7 driver disk that shipped with each Getac and push them onto the image. Most of our current devices were using factory drivers. At the time of imaging them, this was all that was available. Our cards were running firmware version 1.08.16.05. I didn’t see the problem, because when I first imaged the devices, I had no idea how to check the firmware, since there was only one package at the time. On top of that, we have since downloaded, and re-installed the newer driver to the devices. Note: Our Getacs are using the EM7355 cards.

Well after a lot of frustration, I found out what was going on. Even upon installing the newer driver, the firmware was still not updating. I would download and install the 05.05.16.00, and the firmware on Skylight would still show 1.08.16.05. Now, OK, Getac did document this… well sorta. On the download link, you will see the message (only for 4G module firmware version 05.05.16.00 for Skylight. See below:

skylight2

 

Ok, well now the firmware on my device still shows 1.08.16.05, and I need 05.05.16.00. How do I do that? I have downloaded it, and installed it, and it is not upgrading the card firmware. While Skylight does still operate while using card firmware 1.08.16.05, it does not work well… Here is how I found out how to upgrade the firmware.

Now, this may be an easier way to upgrade the firmware, but this is the only way I have figured it out so far. That is to remove the device and driver from the system and install the latest driver. Now, it is finals week, and it is late, and I should be studying, but I am writing this instead, because I would rather do this, instead of re-watching 30 minute lectures on binary search trees. I do not have access to a Getac right now, so I can’t get any screenshots. I will just have to explain it in words.

First off, pop open control panel and remove any software associated with Sierra Wireless. For us it was the driver and Skylight. Once removed, I went to Control Panel > Administrative Tools > Computer Management > Device Manger. From there, expand the network devices and find the 4G card (ie: Sierra crappy 4G card thingy) and select uninstall. MAKE SURE TO CHECK THE BOX AND REMOVE THE DRIVER TOO. This will ensure that we get the files for 1.08.16.05 out of the system.

Edit: Turns out you only need to install the new package over the existing one, then upgrade it by going into Skylight and selecting your carrier, and it should auto upgrade the firmware. When you install the package, you get nothing. It installs the new drivers, but the firmware is not yet upgraded. Proceed by going into Skylight settings, and firmware tab and select a carrier. The firmware upgrade should start. If you already have your carrier selected, then you will need to select another carrier, upgrade it, and then select the new one again. Stupid way to do it, but it works.

Spoiler: I am going to jump ahead here. We got the card firmware upgraded, to 05.05.16.00 and it still had the same message. Searching…. The only other idea I had was to upgrade the Getac Utility that was recently upgraded, in some magical attempt that it would do something. Well it did. Once it was updated, the 4G card connected right up to Verizon, and has been happy ever since. We have also only replicated this on machine so far, since we found the issue out when I was off at College. During my winter break, I will be updating the remaining officers Getac’s to see if makes a difference. Fixing issues or not, this post is about upgrading firmware, so yeah.

Now, it is time to upgrade the Getac Utility. Download the latest one from Getac’s site:

utility

 

Install it, and proceed to install the 05.05.16.00 firmware from Getac’s site:

firmware

 

Once the firmware is installed and complete, install Skylight.

And once that is done, make sure your configurations in Skylight are correct.

We use Verizon, so that is our carrier we have selected. Just for giggles, select it again from the drop down menu, even if it was already selected, and see if it changes the firmware again. You should already be up to date, but if not give that a shot.

Also, we have our profiles connect automatically on startup, and connect even when roaming. That is just our priority as a police department.

That should be pretty much it. Make sure you reboot also to test it. I haven’t had much time to play with the new Getac Utility, but it offers some cool options for passthrough while docked, and more fine tuned options.

ALSO! I just remembered, under the ECO tab in the utility, we have started selecting the WWAN to be set to on. By default it is off, but the card still works. We don’t know what it does, but it can’t hurt right?

Thanks for reading, and like always, if you have any questions or comments regarding this post, please email me at gbrewster@agoasite.com

Finding the proper profile solution

Finding the proper profile solution

When working in a Windows and Citrix environment, things can get a little complicated. It’s not super complicated, but just enough when a simple setting change makes all the difference in when a profile is copied locally or to a remote location. Allow me to explain using my history of profiles.

Up until virtual desktops were deployed at work, all profiles were local. Anyone who logged onto another machine, had no files, and no appdata. This was perfectly fine since we had the idea that each person pretty much had their own computer. Well, introduce XenDesktop 5, with no PvD and randomly pooled desktops. In this case, you must enable folder redirection or use Citrix Profile Manager. After some issues, we went with folder redirection via GPO. This worked fine for the most part, except everything was redirected. EVERYTHING. This caused conflicts with appdata and link folders when applications installed stuff to them. Now after this had been in place for a year or so, I come into the picture and decided to rebuild the virtual environment from the ground up. New folder redirection scheme, new file server, etc. I introduced folder redirection as well. Only I redirected only part of the profile

  • Desktop (Because local government employees like sticking large files on the desktop instead of an organized folder)
  • Documents
    • Pictures
    • Music
    • Videos
  • Contacts
  • Favorites

That was it. The reasoning behind no redirecting appdata was simple. We are small enough, we do not need to do this, and it can and will conflict with other computers down the road. It is nice because a lot of data transfers nicely, but has the potential to be disastrous.

Anyone who has configured folder redirection knows that is user based. I knew that when I first set it up, and it caused me some grief. Our AD environment was setup to be organized as best as possible. Servers, physical machines, virtual machines, and employees all got their own OU’s that allowed me to assign proper policies on a very detailed level. The employee structure when something along the lines of building->department->user sections. Since our structure as a city was set up based on buildings. we don’t have accountants in the police department and public works. They are in one location, which makes our structure very easy to set up. So under each department(if configured with virtual desktops) was a OU for folder redirection. This effectively split up our users. No problem really, they are already organized under their department OU. The problem is when a user with folder redirection logs onto another computer. For the most part, their profile is directed just fine, and it works. In cases with laptops, and computers with slow connections, such as VPN’s, it makes management a nightmare. So I went ahead and tried to find a solution, which involved pulling a WMI request for if the computer was a server (for worker servers) or a laptop. If it was, overwrite their policy for redirected profile and assign a local one. This pretty much double the login time. We kept this policy in place for a number of months until recently I have been getting fed up with some weird issue that kept copying my redirected profile to a local one, and removing it from the server. I finally began to research other options.

Well I found something that works perfectly for our company. Folder redirection based on computers. Now, before you ask yourself, isn’t this what he did before? Well… it kind of is, but this one is a lot better. I’ll explain.

Using the power of GPO loopback, I can assign user options to anyone who logs onto a computer. Can you see where I am going from here? Basically, if any user logs onto a computer that has folder redirection enabled (App servers and virtual desktops) then give them a redirected profile. While this does now make two profiles in some cases, it is way easier to manage than hunting down a lost profile that has been copied to a local machine. In addition, over time, desktops may have folder redirection enabled by default. This will allow for even better integration. But for the time being, it allows me to have tight control on which machines can get redirection and which ones cannot. obviously Windows tablets and laptops need local profiles (because I do not like offline files) and simply need to be in an OU that does not allow folder redirection. What makes it better is when John Doe wants to use Suzanne’s laptop, he won’t get redirected either.

All in all, this choice has helped me un-clutter my AD environment and improve login times. It gives me a better degree of control on when profiles are redirected and when they are not. While this solution is not for everyone, it works for me, and I think we will continue to use it.

5 things I would like to see improved in XenCenter/XenServer

5 things I would like to see improved in XenCenter/XenServer

I have been working with XenServer for about three years now, and I came into the game around 5.5, 5.6, and struggled through the 6.0 conversion. I have been running 6.2 for close to a year now, and It is pretty dang good. We don’t have memory leaks, random vm hangs, and it is just all around better. I only expect XenServer to become better and better. In saying so, I hope that they improve some features that are already implmented. I present five things I would like to see changed or improved in XenServer/XenCenter.

1) Easier XenTools upgrading

XenServer 6.0, 6.1, 6.2. Each upgrade required a XenTools upgrade on each Windows machine, and required each one to reboot. Sometimes more than once. XenTools is quickly becoming a hassle to deal with. You can’t ignore it, because then you won’t get any good performance out of your VM, and it requires a reboot, which often leaves you working late into the night. I know VMware has a one click VMtools upgrade for some of its supported OS’s. I would like to see something similar for XenTools. Now, I am not a software developer for Citrix, so I don’t quite know if such a thing is even possible, but a button that would send commands to a Linux or Windows and perform a silent install, then place an alert XenCenter once it is done, and ready for a reboot. That way a reboot can be scheduled when necessary.

 

2) Faster console refresh

I have never enjoyed using the XenCenter console. It is slow, small, and while it does allow for pasting and copying, it does not refresh super well. From my understanding, it uses VNC to show the screen, and  I have never had good luck with VNC. It never refreshes quickly, and it is just usually a pain to navigate. Not to praise VMware, it has its issues too, but I never had a problem using their console, except the whole mouse release thing. I hope to see improvements in video refresh and performance. Something where I can just log into XenCenter instead of SSHing or RDPing into a server, just to get a better experience.

 

3) More space reserved for XenServer installs

XenServer install size is under 4GB, and that includes the database of VM’s. The rest of the space is used for local storage. If anyone has ever updated XenServer, you have come across alerts that mention that your partition is getting full. This is usually due to log file sizes, which at times reach ~500MB of text files that have not been rotated out. I guess the title should instead be called, ‘option to turn off debugging level logging’, or log file clear button or option.

 

4) VM clones that use specified vm name as HDD label

I cloned a few VM’s, and named them all differently, and then checked the SR to see their names, to find they all had the cloned name. Now while you can change this during the VM create wizard, why is it not done by default, like when you create a new VM from scratch? While everything can be changed later, it just seems more efficient to me to have everything named the same upon creation.

 

5) Remote Desktop using XenServer login credentials

We already authenticate with AD for users in managing XenServer, why can’t we use the credentials to authenticate with RDP to log into servers as well. Having to type in my password each time I already log into a server, and then have the RDP service poll and try to switch ten section after you log in can get annoying. It often requires me to type my password twice. Something with the option to do so, would make my day just a little easier when managing servers.

Now, I don’t know if any of these ideas are even possible to implement, but boy would it be nice to makes things just a little bit easier in XenCenter. I do like the program, for the most part, and I hope to see more options implemented later, as I think XenServer has the potential to become a much better and popular hypervisor.

SandyNet Fiber – First Customer

SandyNet Fiber – First Customer

Yesterday, August 27th, 2014, The SandyNet team rolled out to our new subdivision, Snowberry, to hook up the first fiber to the home customer. After months of preparation, we finally received our equipment, got our data center built, and learned how to splice. Yesterday,we put it to the test, and attempted to get our first customer online. Joseph Kemp was our first customer to obtain SandyNet Fiber to his home. IMG_20140827_151625335_HDR

 

Upon arriving, we pulled open one of our vaults. From here, we began to drag our drop cable to the house.

IMG_20140827_152546748

Each house in Snowberry has a smart panel, which makes it easy for us to get the fiber into the garage, with minimal effort. You can see that phone, Ethernet and coax are all already present.

IMG_20140827_151843853_HDR IMG_20140827_151751311_HDR

 

We then connected the house to the pedestal near the vault. We also had Carlos, being Carlos.

IMG_20140827_153549778_HDR

We finally got to our our splicers in the field. I spliced one pigtail in the garage, and Carlos Spliced the outside pigtail on the house.

IMG_20140827_154558571 IMG_20140827_154856823

Finally got to mounting the ONT and providing it with power. This process took a bit longer than we expected, as we soon found out half of Snowberry was not lit up. Using some basic troubleshooting, we soon discovered that it was not our splice job, or our fiber. We had to call out our contractor (Route 26 Fiber) to take a look and see if anything was missing.

8/28/14 – Thursday Morning

After about an hour of troubleshooting, we shot a visible light beam down one of our strands, and discovered that  our colors in one of our splice cases was incorrect. The problem was quickly remedied and the customer came right online.

IMG_20140828_094643388 IMG_20140828_094632734_HDR

So now, we have technically turned on our first SandyNet Fiber customer. It has been awesome to get involved, and to be apart of the SandyNet team. And while I go off to school soon, I have gotten to be a part of an awesome project. 😀

My time at SandyNet and a brief history of SandyNet Fiber

My time at SandyNet and a brief history of SandyNet Fiber

Note: This post was was written at 2am, and probably contains many spelling and grammar errors.

I have been with the SandyNet team for over five years now, and even when I went off to college, I was asked to work remotely and take calls after hours. It has been an awesome experience, and I am very lucky to still continue to learn and expand my knowledge by working with this company.

SandyNet has changed a lot over the past five years that I have worked there. What was originally started by contractors, was taken over by Scott Brown whom left and was replaced by Joe Knapp. After Joe has been in place for a few years, I was introduced as an Intern. What started out as organizing parts, prepping Ubiquiti legacy nanostations and best of all, making 50’ Ethernet cables, turned into a full blown job, where I had responsibility and I could rise as high as I wanted. Going back, I was absorbed by the city after spending a year volunteering as the IT guy for the Sandy Police Department. I knew basics about networks, and mainly focused on physical computers. Learning how the CPU works, and how memory is allocated. It was 2009 and I was finishing my freshman year of high school and began working every day after school.  Anyway, I began to learn the basics of wireless and got a better understanding of routing, switching and TCP/IP. After a month or two, a new employee was brought on, Carlos Manzano. I grew interested in datacenters, servers and Microsoft Windows. I dinked around with servers and tore up my house dragging Ethernet. About a year later, I obtained my license, and was able to start repairing and setting up customer equipment. Soon enough, a Cisco IP phone showed up on my desk. I remember excitingly telling my parents that I got a phone and was earning more trust in the company. This opportunity came at a price. Helpdesk. The worst job at an ISP. What this boils down to is, you only get a call when something isn’t working. This ensures that the customer is usually in a bad mood, which means that is my fault that they cannot play Call of Duty. Remember, we are a wireless ISP, with three people. We have never had the man power to install every customer with a wireless radio. We usually had the customer pick up the equipment at city hall and install it themselves. Because the customer does not know how a point to multipoint wireless network works, it usually means the equipment was installed incorrectly, resulting in poor performance.

Come late 2010, I started to attend the Center for Advanced Learning. A charter school in Gresham that allowed me to focus on computer related learning. I spent half the day at CAL and the other half at Sandy High School. It was at CAL where I began studying and learning the Cisco CCNA course. In addition to networking, I was introduced to programming in PHP and Java. I also learned how to install, operate, distribute and manage Linux operating systems. Later we moved to the Microsoft learning academy, where I was properly learning how to operate Active Directory environments, DNS, DHCP and other services that Windows server offered. I used this knowledge to set up my own AD environment at home, which later helped me get into managing our internal network at the City of Sandy. By now, Scott Brown has made his way back to the City and we expanded to four employees. Also, at this time, the customer count was also rapidly increasing towards 1000 customers. Now, Sandy has a population of around 11,000, with about 3600 houses. Four people had taken almost 1/3 of the internet customer base in the whole city. I continued to rise slowly, taking on more responsibility. I began to host servers out of the City, which I used to learn more Linux and how to host websites. How DNS really worked and got to deal with security and learning how to properly secure data and sites. I started a little business, where I could host websites and manage them. I mean, we had the bandwidth and power in our datacenter, so why not?

After graduating high school in 2012, I moved onto my first year of College at Mount Hood Community College. I believe CAL helped prepare me for college and made the transition fairly smooth. I also noticed I was changing as well. I felt I needed to be more professional. I had spent two years learning how to talk with people on the phone, and deal with them in person, but I lacked the professionalism. Well, I felt I wanted more of a challenge, and began to ask to start managing servers and computers internally. I started with the deployment of Sophos antivirus to about 80 machines city wide. After that came active directory and managing our Citrix environment. XenApp, XenDesktop, XenServer and Netscaler. Learning to manage and run these systems made the main internal guy. Any problem with a server, or a client device was now my problem. I moved less and less away from phones, and focused more on data, servers and system administration. Currently, it is what I am still doing. After spending a year at OSU(Second year at college) working over a VPN and phone, my main focus was no longer going out and fixing customers problems, but rather me fixing logical, or system problems. This has lead me to spend less time on the SandyNet side, and more on the City of Sandy side. Both sides have benefits and detriments. My most recent task was deploying a new tablet solution to replace our aging CF-29 toughbooks, along with migrating storage data to another server. Fun…

Beginning to get burnt out on system administration, I wanted to get all the piss and vinegar out of my veins before was confined to a desk and phone and left to weep in a corner as stress and boredom got the best of me. I wanted to still have some ties to SandyNet, because it was where we were all trying to go. It was department that was changing all the time. New problems, new solutions. I decided to blend the two. I asked to be put through tower training class, and bucket truck training. This would enable me to rise from being a paid intern, and become an official City of Sandy employee. It took four years, but my responsibility rose from making cables, to managing all internal servers, switches, systems and users. To climbing towers, and replacing broken access points at night. To deploying and ensuring police officers have a system that they can use without worrying about possible problems.

With a department of four people, we have no predefined roles. I am not limited to helpdesk. I am not limited to internal city administration. I am free to move around and show interest in other sections. Because of where I work, there is no written way to do your job. There is no written solution. When a problem shows up, it is my, and everyone else’s job to come up a solution that best suites our needs. The factors of cost, effort, urgency, efficiency and management all play a role in problem solving at SandyNet. We are constantly changing, and we are all moving around from task to task. A project I started may be handed off, or later fixed by my coworker, but we are all in the same boat. We all know the same basic stuff, and all employees can almost pick up any project and work with it. I believe this is one of the main reasons SandyNet has been as successful as it has.

Currently SandyNet is still mainly wireless. Our traffic continues to grow higher and higher, and our equipment continues to fail and buckle under heavy loads. But we have a solution in place, and it is going to solve all of our problems. Fiber.

The whole SandyNet team has been crying for fiber for the past three years. We saw that internet was quickly becoming a necessity and people needed more of it. We knew wireless had limitations, and we had reached our carrying capacity. If SandyNet wanted to have any future at all, it needed a new medium on which it can deliver service. We all decided that fiber was needed.

Well the first two attempts were complete flops. We tried to do it ourselves. We tried to get it in order, and we tried find money to support the project. It was a huge gamble, and nobody would open up. We quickly realized that we needed the help of another company. So we sought out i3 America. A UK based company which held a patent on a fiber technology that was capable of being placed in sewer or water lines. As a local government agency, we were looking for the least amount of production cost, and we saw this as a way to decrease the cost to implement fiber. Things stated moving along. We got in touch, and plans were being drawn out. A few problems did occur. Based on our numbers, we would need 50% take rate from the city for the project to be successful. This was quite a high number. If we could not meet that take rate, we would flop, and SandyNet would most likely go under. I mean, with a project that costs over 8 million dollars, you can’t just hope for the best. So we started advertising. Presenting to people that we have a solution for the stupid slow internet in Sandy, and we need your help to get it going. Orders flowed in. We were getting a lot of people, but it still was not enough. The gamble was that we were going to go for it. So plans for constructions were being drawn out by engineers. Then one day, all communications stopped. They stopped responding to emails and phone calls. No one could get ahold of them. We knew something was wrong. We began to look for a way out, a loophole in the contract. We had a deadline, which was rapidly approaching, and they were not going to meet it unless they got started soon. The deadline was reached, and we still did not have any communication.

Communication was finally re-established. The company has been bought out and was under reconstruction. The UK company went from i3 to SCIFI networks, or something along those lines. We didn’t really care, we just wanted fiber. We began from scratch again. Starting with contracts, the long process of coming to an agreement was difficult. That is about as far as we made it with that company. They could not provide a good enough agreement for us to sign off on. So we dropped them, and went with another company, OFS, which at this time was able to do a traditional fiber build for an even lower price. No mucking around with sewers now, it can be straight in the ground! We quickly jumped on board, and things started rolling. Now, I was in Corvallis when everything started getting back on track. Contracts were signed, plans were done, and equipment was arriving. I came back to have my cubical trashed and filled with modems and fiber from OFS.

Construction had official started a week before I made it back home. It was extremely exciting to see, as fiber was actually going in the ground. It was no longer just talk, or this or that, I could go out and see it with my eyes. One major project upon my return was getting a new datacenter built and prepped for the fiber deployment. Construction on our old office at City Hall was underway while I was in Corvallis, and electrical equipment was installed, and Chatsworth products arrived three weeks into my return for the summer. I then spend another three weeks assembling racks, hanging ladder systems, dragging cable, fiber and mounting racks in the ground. PDU’s were in every rack, and two Star bus systems were in place to distribute 90% of the power we used in the room. New Cummins generator, and fancy Eaton battery backup system. It was extremely awesome to actually build a datacenter. While I did not do electrical or construction the room itself. I did assemble and mount all equipment inside. Once it was all done, we performed our generator and battery tests and spent two nights, a week apart, migrating all equipment from one room to the other. With my two hands, I had actually mounted $250,000 worth of Calix equipment in a matter of hours. I set up, blew up and then helped repair a DC plant, and wired all Calix gear to it. It was an awesome experience.

Now, I have had to leave out so many details about this project, and my history with SandyNet. There is enough experience and memories within these five years to write a book. I am very glad to have this job, and to continue working and rising up in the company. It is ever changing, and by staying here, I am getting awesome experience in one job, that five normal jobs could not even begin to scratch the surface of.

If you speak to our customers, you will get both good and bad information. Yes, wireless sucks. For all that is good and holy, stop reminding me. I work with it every day, I know your pains. But all is not lost! Fiber is around the corner. Hopefully by the end of 2014, fiber will in the ground and the city will be deployed. The process of hooking people up then finding out what to do with all the money that will be pouring in.

I fear that after the fiber project, things will get boring. I hope this is not the case, but having spent so much time with the city, it is hard to say what exactly will happen. Only time will tell. And if I do not like where it is going, I can always quit and go somewhere else. Either way, the experience alone is a story I can tell for the rest of my life. Not a whole lot of people get to experience what I get to. I am grateful to have this opportunity, and to be a part of this small time ISP. As of right now, 8/16/14, there are not a lot of municipal ISP’s that have fiber. I am excited to see how this goes, and how I can keep growing and learning new things. I have been gifted with the want to continue learning. To want to learn more, and to keep moving and not get bored. I hope to be able to brag later in my life, that four people, ran an entire city’s internet.

Next week, we are hoping to get our Calix gear provisioned and ready to roll out. If we are lucky, we will be turning on our first customers in the Snowberry subdivision. I hope to be the guy that turns on the first fiber customer in the small city that is Sandy.  😀

I got a few photos of our data center build. We had contractors come in and smash a couple of walls, then build a few walls, put in AC, and wire up an automatic transfer switch and install a star bus power line. My job was to unbox, assemble, and mount all the Chats Worth hardware that arrived. Which there was quite a bit of.

 

IMG_20140624_140753809 IMG_20140624_141057696 IMG_20140625_132512699

 

 

 

 

 

Once everything was unpacked, we began install the ladders and racks. We spent a lot of money to get some good cable management so that everything would look nice. We barely used any Velcro or zip ties to secure wires which is awesome.

IMG_20140701_211136601 IMG_20140701_183243922 IMG_20140701_211149246 IMG_20140703_180455525 IMG_20140709_150917379

 

 

 

 

 

 

 

 

 

The next task was to get Ethernet to each rack. Each of our three 4-post racks had 48 strands of Cat 6 connecting back to a two post rack for distribution.

IMG_20140703_180509467

IMG_20140709_114938124 IMG_20140709_114949960

 

 

 

 

 

Then the electrical was finished up, and we got all of the racks mounted with PDU’s and connected to the Star Line power bus.

IMG_20140709_151046727 IMG_20140709_165722929

 

 

 

 

 

The Calix gear was mounted and ready to be turned on. At this point we had to wait a week or so until the DC plant arrived.

IMG_20140703_171156nopm IMG_20140703_180448574_HDR IMG_20140705_161457582 IMG_20140705_161527734 IMG_20140705_161834589 IMG_20140709_114956189 IMG_20140818_100651435 IMG_20140818_100639876

 

 

 

 

 

 

 

 

 

 

 

 

 

After blowing up three rectifiers, we finally got all of our DC equipment setup and things were running. We then migrated all of our existing hardware from our old server room to our new one.

IMG_20140819_160419620_HDR

 

 

 

 

 

On 8/19/14 we got our first ONT setup in my truck, and tested the Calix gear to make sure it works. Soon enough we will be able to start hooking up customers, and making people happy!

So we went from this

IMG_20130924_140330910_HDR

to this

IMG_20140819_171557915_HDR

and now the old server room looks like this

IMG_20140819_171704029

We’ll clean it up someday. I promise.

 

Powered Parachute Training

Powered Parachute Training

This last weekend, 7/26/14 I have the opportunity to get some training on how to operate and fly paraplanes. An awesome experience taught by Mark Martin. This post is just a footnote, instead of a lengthy post about what I did. I have a video and few pictures of my experience. My father and I are setup, and planning on obtaining our sports pilot licenses. With that, we will be able to fly a passenger when operating a two person power chute. A lot more work to come to obtain it, but we are hoping it is something we will enjoy and continue to do.

 

Getac F110 Imaging

Getac F110 Imaging

Recently at work, we obtained 21 new tablets to replace the current CF-29 Toughbooks that were in production. It is currently my job to set up the Getac F110’s. The process has been… Painful.

While I was finally able to finally get a Getac to image, I was unable to get it to image through the docking station NIC. Instead I found myself buying a USB to Ethernet adapter. While I am sure there is a proper way to image these devices, I have contacted Getac support a few times, and they either never replied to emails, and on the phone I got the usual run around for possible fixes. Let me explain the scenario.

We are deploying Windows 7. Because we do not have time, and it is not our jobs to hold employee’s hands as they learn a new operating system UI. With that said, we got our Getac’s shipped with Windows 7. I had previously set up a WDS server for imaging on our network running Windows server 2008 R2. I will attempt to explain the process to the best my abilities. I understand this is not a solution, but rather a workaround. Because Getac does not have much documentation available on their site, I am stuck with getting machines out on a deadline, so I don’t have time to do it the proper way.

Crap you will be needing:

  • 3 flash drives (Yes, 1 for BIOS, KBC BIOS, Window PE image)
  • Getac F110
  • F110 Docking Station
  • Rufus USB program
  • Alcohol

Note: Getac F110’s are purely UEFI. With this said, it does have a legacy mode, but it does not work with the NIC. The only legacy devices you may boot from are the SSD, and USB drives. PXE boot is strictly UEFI, and there is no legacy mode, which causes me a lot of pain when trying to boot them from a PXE server. The work around has been to use USB drives and boot into those to download an image from WDS or MDT.

So it is always a good idea to update to the latest BIOS before you ship the computer out to production. The current BIOS that are shipped on Getacs (at least ours) does not offer PXE Boot. By updating to it, it was added and enabled, but I was not able to get it working.

  1. Get it hooked up.
    1. I used a USB to Ethernet adapter for initial setup, because I was trying to solely use WDS to deploy and not MDT. After a lot of blood, sweat and tears, I moved on to MDT for deployment, so I am not sure if a USB to Ethernet adapter is still needed, but it doesn’t hurt!
  2. Update the BIO’s
    1. Download the proper BIOS zip file from Getac’s site
    2. Using Rufus, format as FAT and MSDOS, so it is bootable
    3. Extract and copy all of the contents onto the newly formatted USB drive
    4. Plug the USB drive into the Getac and boot into it
    5. Update the BIOS
  3. Update the KBC BIOS
    1. Download the proper KBC BIOS zip from Getac’s site
    2. Using Rufus, format as FAT and MSDOS, so it is bootable
    3. Extract and copy all of the contents onto the newly formatted USB drive
    4. Plug the USB drive into the Getac and boot into it
    5. Update the KBC BIOS
  4. Create boot image – Since we are deploying Windows 7, We will need to use a Windows 7 boot image. Start by fetching a Windows 7 installation disk.
    1. Import the boot.wim file from the installation disk into your deployment server.
  5. Install MDT (Microsoft Deployment Toolkit)
  6. Create a project in MDT called Getac or something

    1. Set the deployment share
    2. Do not ask if a computer backup should be performed
    3. Do not ask for a product key
    4. Do not ask to set the local admin password
    5. Ask if an image should be capturedshare
  7. Take the Getac Windows 7 Driver disk and copy it over
    to the server. There should be a crapload of drivers so it will take a while

    1. Now, back in MDT expand Out-of-Box Drivers and right click on it and select Import Drivers
      1. Browse to the driver location and select the file containing all of the driversdrivers
    2. This will take a long time
  8. Right click on your deployment share and select properties
    1. Under the General tab, remove the x86 platform under ‘Platforms Supported’
    2. Under Windows PE, select the x64 Platform
      1. Under the features tab, add the DISM Cmdlets
      2. Under the Drivers and Patches make sure the Include all drivers from the selection profile is selected
      3. Do not forget to add the inf driver files for the USB to Ethernet adapter
    3. Close out of the properties window
  9. Under the Operating Systems tab, add a new OS, and navigate to the install.wim file from the Windows 7 install CD you copied earlier. If you would like, you may add images from a WDS server(which I did to save space) but it is not required.
  10. Lets make some tasks, right click on Task Sequence to make a new task
    1. Give it an ID of like 1337 or 9001 or something and name with the words capture in it.
      1. Under Select Template make sure Sysprep and Capture is selected
      2. Select the Windows 7 Pro OS from the selection
      3. Fill out the rest of the information if you wish
      4. Finish it and it will now be present
    2. We will make another template for deployment once we have captured our first image
  11. Fun time! Take your base Getac and make an image for it. Install updates and whatever software you wish. Once you are done, do not sysprep the machine. Just unjoin it from a domain if it is currently connected to one.
  12. Capture the image and upload it to the deployment share capture location. (For capturing this image, I used a guide found here)
    1. Basicly, on the golden image or base machine, we are going to navigate to the network share that we set up earlier when we created the deployment share.
    2. Select and execute the LiteTouch.vbs script to start the capturing processscripts
    3. When it is done, it should show up under the deployment share/captures using the ID we specified for the capture template
  13. Now we have captured the base image. YAY. If you are working along side with WDS, upload the captured image to the WDS server. Then add all new WDS images to the deployment share. If you are not, just upload the image to the OS’s in the deployment share
  14. Now go back to task sequences and make a new one for deployment
    1. Select Standard Client Task Sequence
    2. Select the newly captured image for deployment
    3. Fill out the rest of the settings as you wish
    4. Hit finish, and the deployment task should now be present
  15. Make a bootable USB drive using the TouchLite PE.iso image
    1. Select your deployment share and update it. This will take a long time because we need to inject all of the drivers. Once it has completed, we can find the iso and write it to a USB drive.
    2. Locate the file under \\server\deploymentshareaddress$\boot\
    3. Using Rufus, write the LiteTouchPE_64.iso to the USB drive
  16. Plug the USB dongle into the Getac’s one USB port
  17. Dock the Getac and turn it on. Hold the Windows button to access the boot menu
  18. Boot from the USB dongle
  19. Once the LiteTouch program and Selected Deploy when the button appears
    1. Select the deploy task from list of tasks.
    2. Fill out the information and the process should somewhat automate
    3. Once the image has deployed, it will log into it, and attempt to contact the server again. For this to take place, make sure there is a cable connected to the docking station as well as the USB to Ethernet adapter.
    4. You should receive a yellow box saying it was completed.
    5. Unplug USB to Ethernet adapter.
  20. You have just imaged at Getac F110!

I hope others find this helpful, if anyone has any questions, let me know at gbrewster@agoasite.com

Certified for Tower Climbing and Rescue

Certified for Tower Climbing and Rescue

I have been busy with School. And Work. And other things. But, I recently took two days off from school (bad idea) and went to a training session in Clackamas by Gravitec for getting certified to climb towers. Attending with Legacy Wireless, I spent two full days learning to do’s and don’ts. What type of safety equipment is needed, and how to use it. Both in the classroom and out on a small tower, we practiced rescue and climbing techniques that we can use in case of emergency. Once I was finished, My coworker took me out to one of our sites and I climbed my first full sized tower. Of course I wore my Glass, and took some pictures to document this achievement.

 

Doge App

Doge App

When I have free time, I have been slowly working on an Android live wallpaper. While it is very broken and worthless right now, I plan to eventually have it do something. Right now, it generates random doges and they scroll up the screen.

It is still very much a work in progress, but plan to see it on Google Play someday. I hope.

Screenshot_2013-10-31-17-47-25