First, getting headless.

I will not be using the "Raspberry Pi" as "a common PC"; there are already too many of these at home. And well, for the use-case I am thinking of, it is not needed. So the first step is to make it "headless".

For my initial experiment, I took the default Linux distribution "Raspbian wheezy" (16/12/2012); I preferred to start with the "official" distribution (I assume, the most stable and the easiest to begin with). 
I made sure I would have at least:
  • A connection to the screen: that meant for me a cable HDMI to DVI.
  • A network connection to my home modem.
  • A USB keyboard.
Everything went well, except that the keyboard I used (a Logitech Wireless keyboard K340) did not work out of the box. For some obscure reason, if the USB dongle is connected at boot-time, I don't get the keyboard to work; but, if I unplug and plug again the dongle after the RPI booted, it worked just fine...

After that I could finally start. Good thing is that the configuration tool ("raspi-config") has the menu for enabling the SSH server. So after changing the default user password, extending the root partition to fill the SD-card (I bought a 8G) and enabling the SSH daemon, I was almost ready.
Apparently by default there is no password assigned to the "root" user; but well, for safety, I preferred to disable SSH root login, just in case. This is done by editing the SSH daemon configuration file "/etc/ssh/sshd_config" and make sure that "PermitRootLogin" is set to "no". Voila, at the next restart I will be ok.

The final step to make the RPI headless is to ensure it will always get the same IP address. There are two simple solutions in my case:
  1. Make sure my home modem assigns a static lease IP address for the RPI (based on the network MAC address).
  2. Configure the RPI network to use a fixed IP address.
Unfortunately, 1) failed badly... :-( The Belgacom "BBOX" modem lost the DHCP static lease bindings at the next reboot of the box :-( So for now, 2) is the way to go.

This is done by editing "/etc/network/interfaces" and making sure to write the following content

iface eth0 inet static
        address 192.168.1.100
        netmask 255.255.255.0
        gateway 192.168.1.1

Now, my RPI is ready. I don't need the screen and the keyboard. All it needs is USB power and network. Connected to the USB port of the modem, it boots when the modem is up and I can remotely connect to it via SSH, everywhere... from home, at least for now.

Comments

Post a Comment

Popular posts from this blog

ZWay: getting the device list.

Back after a while... and a son! Here is a piece of code that can be used with the latest version of the ZWay library. It makes it possible to retrieve the Z-Wave device list and display the device names and associated XML file (each recognized device can be associated to an XML that contains a bunch of meta-data: name, description etc.). int main(int argc, char ** argv) { ZWay theZWay; ZWError result; memset(&theZWay, 0, sizeof(theZWay)); result = zway_init(&theZWay, "/dev/ttyAMA0", "/home/pi/src/razberry-proxy/config", "/home/pi/src/razberry-proxy/translations", "/home/pi/src/razberry-proxy/ZDDX", stdout, Warning); if (result == NoError) { result = zway_start(theZWay, &terminationCb); if (result == NoError) { printf("Discovering...\n"); result = zway_discover(theZW...
Read more

First steps with the ZWay library

The Z-Wave "RaZberry" installation comes with everything needed to be able to do some custom development in C; everything is installed in the /opt/z-way-server  directory. There, one can find: A set of (documented, nice!) C header files found in /opt/z-way-server/libzway-dev : CommandClassesPublic.h FunctionClassesPublic.h ZDataPublic.h ZDefsPublic.h ZErrors.h ZPlatform.h ZWayLib.h A set of libraries found in /opt/z-way-server/libs : libmicrohttpd.so libv8.so libzwayhttp.so libzwayjs.so libzway.so For the ZWay stack to run properly, it is needed to provide the necessary configuration files. Instead of using these directly from the installation folder, I made a copy in order not to mess up too much with my working/configured setup. The following directories from the ZWay installation folder are needed: config translations ZDDX For experimenting I made a small Makefile in order to build and run. main.out: main.c gcc -I/opt/z-way-server/li...
Read more

Libcurl: perform a REST HTTP PUT

The libcurl library does not provide a clear example on how to issue HTTP PUT operations in a REST context; the examples are mainly about file transfers . For my proxy application this is not really what I need: I want to perform a REST HTTP PUT, keep it simple, meaning without using the "Expect" and "Transfer-Encoding" that libcurl adds by default. After some struggling I discovered that this behavior can be changed by using the following code: CURL * pCurl; struct curl_slist *headers = NULL; headers = curl_slist_append(headers, "Content-Type: application/json "); headers = curl_slist_append(headers, "Content-Length: 12345 "); headers = curl_slist_append(headers, "Expect:"); headers = curl_slist_append(headers, "Transfer-Encoding:"); pCurl = curl_easy_init(); curl_easy_setopt(pCurl, CURLOPT_VERBOSE, 1); curl_easy_setopt(pCurl, CURLOPT_PUT, 1); curl_easy_setopt(pCurl, CURLOPT_URL, " https://your-server.app...
Read more