Build Backend Pool - Revision history

Agember: /* Configure Xinu Console Daemon */

2010-01-27T03:29:28Z

Configure Xinu Console Daemon

Akoehler: HOWTO:Build Backend Pool moved to Build Backend Pool: no need for HOWTO in the title, its a category

2008-12-19T03:27:26Z

HOWTO:Build Backend Pool moved to Build Backend Pool: no need for HOWTO in the title, its a category

Akoehler at 03:26, 19 December 2008

2008-12-19T03:26:46Z

Amallen: deleted TODO stuff on the bottom. They are now a ticket for someone to deal with

2008-09-25T22:42:40Z

deleted TODO stuff on the bottom. They are now a ticket for someone to deal with

Amallen: /* Mips-Console Wrapper Script */

2008-07-16T19:44:44Z

Mips-Console Wrapper Script

Amallen: /* Console Access */ typos and formatting

2008-07-16T19:32:44Z

Console Access: typos and formatting

Amallen: /* Console Access */

2008-07-16T19:23:17Z

Console Access

Amallen: /* Configure Xinu Console Daemon */

2008-07-16T19:19:48Z

Configure Xinu Console Daemon

Amallen: /* Configure Xinu Console Daemon */

2008-07-16T19:18:00Z

Configure Xinu Console Daemon

Amallen: /* Configure Xinu Console Daemon */

2008-07-16T19:16:32Z

Configure Xinu Console Daemon

@@ Line 108: / Line 108: @@
 <pre>setenv XINU_SERVERS morbius</pre>
-Again, you will want to change "morbius" to match your XINU server's name. The changes to these two files ensures that the default value for the ''XINU_SERVER'' environment variable will be correct when remote users log into the server.
+Again, you will want to change "morbius" to match your XINU server's name. The changes to these two files ensures that the default value for the ''XINU_SERVER'' environment variable will be correct when remote users log into the server. '''NOTE:''' you may need to use the fully qualified domain name of the computer (for example "morbius.mscs.mu.edu") for the Xinu Console Daemon to function correctly.
 Then, open the Xinu Console Daemon configuration file <code> /etc/xinu-consoled.conf</code>. Here is a sample of the configuration file:

← Older revision		Revision as of 03:26, 19 December 2008
Line 1:		Line 1:
		+	[[Category:HOWTO]]
	== Summary ==		== Summary ==
−
	This page details how to scale your laboratory environment up to a pool of backend target machines available for remote access.		This page details how to scale your laboratory environment up to a pool of backend target machines available for remote access.

← Older revision		Revision as of 22:42, 25 September 2008
Line 180:		Line 180:

	Once this change is complete, frontend users can navigate to the directory that contains their <code> xinu.boot </code> image and simply run the command <code> mips-console </code> to receive, connect to, and boot their own modified XINU image on a backend router from the pool using the completely automated script. It will automatically hand out a backend router to the user from the pool and then download their <code> xinu.boot </code> image to the XINU server's <code> /tftpboot </code> directory under the appropriate name for the specific backend router. Then it will run that backend's ''powercycle'' command and send breaks to get to the ''CFE'' prompt. '''NOTE:''' if you do not have a hardware rebooter that can be used to remotely turn off and on the backend, then the frontend user will have to manually restart the router at this step. Next it will automatically configure the backend's IP address using the XINU server's '''DHCP Server Daemon''' and then boot off of the backend's specific boot image located in the <code> /tftpboot </code> directory on the XINU server which should now be the same image that the frontend user developed.		Once this change is complete, frontend users can navigate to the directory that contains their <code> xinu.boot </code> image and simply run the command <code> mips-console </code> to receive, connect to, and boot their own modified XINU image on a backend router from the pool using the completely automated script. It will automatically hand out a backend router to the user from the pool and then download their <code> xinu.boot </code> image to the XINU server's <code> /tftpboot </code> directory under the appropriate name for the specific backend router. Then it will run that backend's ''powercycle'' command and send breaks to get to the ''CFE'' prompt. '''NOTE:''' if you do not have a hardware rebooter that can be used to remotely turn off and on the backend, then the frontend user will have to manually restart the router at this step. Next it will automatically configure the backend's IP address using the XINU server's '''DHCP Server Daemon''' and then boot off of the backend's specific boot image located in the <code> /tftpboot </code> directory on the XINU server which should now be the same image that the frontend user developed.
−
−
−
−
−
−	~~I think these are TODOs : Source for xinu-console included in Console Tools tarball. Explain environment variables. Ssh tunneling? Mips-console wrapper script.~~
−
−	~~== Security ==~~
−
−	~~A word on security. Isolated private network. TCP Wrappers. Iptables packet filtering.~~

@@ Line 177: / Line 177: @@
 === Mips-Console Wrapper Script ===
-So, though it is entirely possible to just use the xinu-console program to connect to a backend and manually send it all the commands to boot XINU, it includes a lot of repetitive actions. So we have included a script for automating the process of booting XINU on a MIPS backend (like our LinkSys routers). This is our <code> mips-console </code> wrapper script located in the file <code> /usr/bin/mips-console</code>. Here is a copy of the [[mips-console|mips-console file]]. In order to get this script to run with your XINU server you will need to modify the top line <code> set ip 192.168.1.2 </code> to match the IP address of the network interface your XINU server uses to connect with your backend pool.
+So, though it is entirely possible to just use the <code> xinu-console </code> program to connect to a backend and manually send it all the commands to boot XINU, it includes a lot of repetitive actions. So we have included a script for automating the process of booting XINU on a MIPS backend (like our LinkSys routers). This is our <code> mips-console </code> wrapper script located in the file <code> /usr/bin/mips-console</code>. Here is a copy of the [[mips-console|mips-console file]]. In order to get this script to run with your XINU server you will need to modify the top line <code> set ip 192.168.1.2 </code> to match the IP address of the network interface your XINU server uses to connect with your backend pool.
-Once this change is complete, frontend users can navigate to the directory that contains their <code> xinu.boot </code> image and simply run the command <code> mips-console </code> to receive, connect to, and boot their own modified XINU image on a backend router from the pool using the completely automated script. It will automatically hand out a backend router to the user from the pool and then download their <code> xinu.boot </code> image to the XINU server's <code> /tftpboot </code> directory under the appropriate name for the specific backend router. Then it will run that backend's ''powercycle'' command and send breaks to get to the ''CFE'' prompt ('''NOTE:''' if you do not have a hardware rebooter that can be used to remotely turn off and on the backend, then the frontend user will have to manually restart the router at this step). Next it will automatically configure the backend's IP address using the XINU server's '''DHCP Server Daemon''' and then boot off of the backend's specific boot image located in the <code> /tftpboot </code> directory on the XINU server which should now be the same image that the frontend user developed.
+Once this change is complete, frontend users can navigate to the directory that contains their <code> xinu.boot </code> image and simply run the command <code> mips-console </code> to receive, connect to, and boot their own modified XINU image on a backend router from the pool using the completely automated script. It will automatically hand out a backend router to the user from the pool and then download their <code> xinu.boot </code> image to the XINU server's <code> /tftpboot </code> directory under the appropriate name for the specific backend router. Then it will run that backend's ''powercycle'' command and send breaks to get to the ''CFE'' prompt. '''NOTE:''' if you do not have a hardware rebooter that can be used to remotely turn off and on the backend, then the frontend user will have to manually restart the router at this step. Next it will automatically configure the backend's IP address using the XINU server's '''DHCP Server Daemon''' and then boot off of the backend's specific boot image located in the <code> /tftpboot </code> directory on the XINU server which should now be the same image that the frontend user developed.

@@ Line 160: / Line 160: @@
 === Console Access ===
-Clients use the xinu-console program to connect their frontend machines to backend routers through the XINU server running the Xinu Console Daemon. Once the Xinu Console Daemon has handed the frontend machine a backend to work on and used '''tty-connect''' to establish a serial connection, the user can use the xinu-console program to interact with the backend by using some built-in commands and also by sending characters over the serial connection by just typing on the keyboard. Here is the ''help'' output for the built-in commands:
+Clients use the <code> xinu-console </code> program to connect their frontend machines to backend routers through the XINU server running the Xinu Console Daemon. To run this program just execute the following command: <code> xinu-console</code>. You may also pass as an argument to this program the name of the specific backend you wish to connect to. Once the Xinu Console Daemon has handed the frontend machine a backend to work on and used '''tty-connect''' to establish a serial connection, the user can use the <code> xinu-console </code> program to interact with the backend by using some built-in commands and also by sending characters over the serial connection by just typing on the keyboard. Here is the ''help'' output for the built-in commands:
 <pre>h, ?    : help message
 b       : send break
@@ Line 173: / Line 173: @@
 q       : quit and power down</pre>
-Notice the key words "''download''", "''powercycle''", "''power on''", and "''power off''" and how they match up with the commands listed in the Xinu Console Daemon configuration file <code> xinu-consoled.conf</code>. Built-in xinu-console commands with these key words in them will call the associated command listed in the Xinu Console Daemon configuration file for that specific backend. To use these commands first the user will have to enter ''command mode'' by pressing Ctrl+Space. The user should see the words <code> (command-mode)</code>, letting the user know that the next character they enter will be interpreted by the xinu-console program as a built-in command and not as just another character to send across the serial connection to the backend. To quit out of the program, enter command mode and then type either "x" or "q".
+Notice the key words "''download''", "''powercycle''", "''power on''", and "''power off''" and how they match up with the commands listed in the Xinu Console Daemon configuration file <code> xinu-consoled.conf</code>. Built-in xinu-console commands with these key words in them will call the associated program listed in the Xinu Console Daemon configuration file for that specific backend. To use these commands first the user will have to enter ''command mode'' by pressing Ctrl+Space. The user should see the words <code> (command-mode)</code>, letting them know that the next character they enter will be interpreted by the xinu-console program as a built-in command and not as just another character to send across the serial connection to the backend. To quit out of the <code> xinu-console </code> program, enter command mode and then type either "x" or "q".
 === Mips-Console Wrapper Script ===

← Older revision	Revision as of 03:27, 19 December 2008
(No difference)

@@ Line 150: / Line 150: @@
 The first line of each entry contains the name of the backend, the architecture it runs on, and the command (along with its arguments) for the server to run in order to connect to the backend. So the line <code> router1 mips /usr/sbin/tty-connect -r 115200 /dev/ttyS0 </code> tells the Xinu Console Daemon that there is a router called "''router1''" that runs a MIPS processor and that to connect to the backend router the daemon should use <code> tty-connect </code> with a baudrate of 115200 on serial device <code> /dev/ttyS0</code>. So, when creating your own configuration file you will want to edit the first line of each entry to reflect the name of each of your backend routers and the serial device connecting it to your XINU server. The other parts of the line should already be correct and should not have to be changed for any of your backend routers.
-The second line of each entry contains the information regarding what actions the server needs to perform to download the specific boot image for that backend router. So the line <code> router1-dl DOWNLOAD   /usr/sbin/cp-download /tftpboot/router1.boot </code> tells the Xinu Console Daemon that the ''DOWNLOAD'' command for "''router1''" will run the command <code> /usr/sbin/cp-download </code> which will take a user specified file on the frontend machine (usually their <code> xinu.boot </code> file) and copy it into the XINU server's <code> /tftpboot </code> directory as the file <code> router1.boot</code>. This lets a student create their own modified <code> xinu.boot </code> image and then, when the server designates a backend for them to use, the Xinu Console Daemon will copy their boot image onto the server with the specific name of the boot image that will always run on that given backend. For your server you will want to edit the second line of each entry so that the name of the connection and the name of the boot image matches the name of that backend router by changing <code> router1-dl </code> to <code> [backend name]-dl </code> and by changing <code> /tftpboot/router1.boot </code> to <code> /tftpboot/[backend name].boot</code>.
+The second line of each entry contains the information regarding what actions the server needs to perform to download the specific boot image for that backend router. So the line <code> router1-dl DOWNLOAD   /usr/sbin/cp-download /tftpboot/router1.boot </code> tells the Xinu Console Daemon that the ''DOWNLOAD'' command for "''router1''" will run the program <code> /usr/sbin/cp-download </code> which will take a user specified file on the frontend machine (usually their <code> xinu.boot </code> file) and copy it into the XINU server's <code> /tftpboot </code> directory as the file <code> router1.boot</code>. This lets a student create their own modified <code> xinu.boot </code> image and then, when the server designates a backend for them to use, the Xinu Console Daemon will copy their boot image onto the server with the specific name of the boot image that will always run on that given backend. For your server you will want to edit the second line of each entry so that the name of the connection and the name of the boot image matches the name of that backend router by changing <code> router1-dl </code> to <code> [backend name]-dl </code> and by changing <code> /tftpboot/router1.boot </code> to <code> /tftpboot/[backend name].boot</code>.
 The third, fourth, and fifth lines of each entry contain information regarding what actions the server needs to perform to powercycle, power off, and power on each router. However, without special hardware to control the power of the backend pool (such as a Serial-Controlled Power Strip) and special software like our Xinu Power Daemon to oversee the use of that hardware these lines will be useless. If you have a more advanced setup with a way to power on and off the backends remotely, then these lines are where you specify the commands used by the Xinu Console Daemon and Xinu Power Daemon to perform those actions. For more information on setting this up, check out our [[Xinu Power Daemon]] wiki. If you are not using any special power control hardware then the user will be responsible for turning on and off the routers by hand.

@@ Line 148: / Line 148: @@
 The line <code> hostname: </code> will need to contain your XINU server's name. Following that line will be groups of configuration information for each of the backends connected to your XINU server.
-The first line of each entry contains the name of the backend, the architecture it runs on, and the command (along with its arguments) for the server to run in order to connect to the backend. So the line <code> router1 mips /usr/sbin/tty-connect -r 115200 /dev/ttyS0 </code> tells the Xinu Console Daemon that there is a router called "''router1''" that runs a MIPS processor and that to connect to the backend router the daemon should use <code> tty-connect </code> with a baudrate of 115200 on serial device <code> /dev/ttyS0</code>. So, when creating your own configuration file you will want to edit the first line of each entry to reflect the name of each of your backend routers and the serial device connecting it to your XINU server. The other parts of the line should already be correct and should not have to be changed for any of your backends.
+The first line of each entry contains the name of the backend, the architecture it runs on, and the command (along with its arguments) for the server to run in order to connect to the backend. So the line <code> router1 mips /usr/sbin/tty-connect -r 115200 /dev/ttyS0 </code> tells the Xinu Console Daemon that there is a router called "''router1''" that runs a MIPS processor and that to connect to the backend router the daemon should use <code> tty-connect </code> with a baudrate of 115200 on serial device <code> /dev/ttyS0</code>. So, when creating your own configuration file you will want to edit the first line of each entry to reflect the name of each of your backend routers and the serial device connecting it to your XINU server. The other parts of the line should already be correct and should not have to be changed for any of your backend routers.
 The second line of each entry contains the information regarding what actions the server needs to perform to download the specific boot image for that backend router. So the line <code> router1-dl DOWNLOAD   /usr/sbin/cp-download /tftpboot/router1.boot </code> tells the Xinu Console Daemon that the ''DOWNLOAD'' command for "''router1''" will run the command <code> /usr/sbin/cp-download </code> which will take a user specified file on the frontend machine (usually their <code> xinu.boot </code> file) and copy it into the XINU server's <code> /tftpboot </code> directory as the file <code> router1.boot</code>. This lets a student create their own modified <code> xinu.boot </code> image and then, when the server designates a backend for them to use, the Xinu Console Daemon will copy their boot image onto the server with the specific name of the boot image that will always run on that given backend. For your server you will want to edit the second line of each entry so that the name of the connection and the name of the boot image matches the name of that backend router by changing <code> router1-dl </code> to <code> [backend name]-dl </code> and by changing <code> /tftpboot/router1.boot </code> to <code> /tftpboot/[backend name].boot</code>.