| 1 |
rcd_serial - RC Delayed Serial |
| 2 |
------------------------------ |
| 3 |
|
| 4 |
This stonith plugin uses one (or both) of the control lines of a serial |
| 5 |
device (on the stonith host) to reboot another host (the stonith'ed host) |
| 6 |
by closing its reset switch. A simple idea with one major problem - any |
| 7 |
glitch which occurs on the serial line of the stonith host can potentially |
| 8 |
cause a reset of the stonith'ed host. Such "glitches" can occur when the |
| 9 |
stonith host is powered up or reset, during BIOS detection of the serial |
| 10 |
ports, when the kernel loads up the serial port driver, etc. |
| 11 |
|
| 12 |
To fix this, you need to introduce a delay between the assertion of the |
| 13 |
control signal on the serial port and the closing of the reset switch. |
| 14 |
Then any glitches will be dissipated. When you really want to do the |
| 15 |
business, you hold the control signal high for a "long time" rather than |
| 16 |
just tickling it "glitch-fashion" by, e.g., using the rcd_serial plugin. |
| 17 |
|
| 18 |
As the name of the plugin suggests, one way to achieve the required delay is |
| 19 |
to use a simple RC circuit and an npn transistor: |
| 20 |
|
| 21 |
|
| 22 |
. . |
| 23 |
RTS . . ----------- +5V |
| 24 |
or ---------- . | |
| 25 |
DTR . | . Rl reset |
| 26 |
. | T1 . | |\logic |
| 27 |
. Rt | ------RWL--------| -------> |
| 28 |
. | b| /c . |/ |
| 29 |
. |---Rb---|/ . |
| 30 |
. | |\ . (m/b wiring typical |
| 31 |
. C | \e . only - YMMV!) |
| 32 |
. | | . |
| 33 |
. | | . |
| 34 |
SG ---------------------------RWG----------- 0V |
| 35 |
. . |
| 36 |
. . stonith'ed host |
| 37 |
stonith host --->.<----- RC circuit ----->.<---- RWL = reset wire live |
| 38 |
(serial port) . . RWG = reset wire ground |
| 39 |
|
| 40 |
|
| 41 |
The characteristic delay (in seconds) is given by the product of Rt (in ohms) |
| 42 |
and C (in Farads). Suitable values for the 4 components of the RC circuit |
| 43 |
above are: |
| 44 |
|
| 45 |
Rt = 20k |
| 46 |
C = 47uF |
| 47 |
Rb = 360k |
| 48 |
T1 = BC108 |
| 49 |
|
| 50 |
which gives a delay of 20 x 10e3 x 47 x 10e-6 = 0.94s. In practice the |
| 51 |
actual delay achieved will depend on the pull-up load resistor Rl if Rl is |
| 52 |
small: for Rl greater than 3k there is no significant dependence but lower |
| 53 |
than this and the delay will increase - to about 1.4s at 1k and 1.9s at 0.5k. |
| 54 |
|
| 55 |
This circuit will work but it is a bit dangerous for the following reasons: |
| 56 |
|
| 57 |
1) If by mistake you open the serial port with minicom (or virtually any |
| 58 |
other piece of software) you will cause a stonith reset ;-(. This is |
| 59 |
because opening the port will by default cause the assertion of both DTR |
| 60 |
and RTS, and a program like minicom will hold them high thenceforth (unless |
| 61 |
and until a receive buffer overflow pulls RTS down). |
| 62 |
|
| 63 |
2) Some motherboards have the property that when held in the reset state, |
| 64 |
all serial outputs are driven high. Thus, if you have the circuit above |
| 65 |
attached to a serial port on such a motherboard, if you were to press the |
| 66 |
(manual) reset switch and hold it in for more than a second or so, you will |
| 67 |
cause a stonith reset of the attached system ;-(. |
| 68 |
|
| 69 |
This problem can be solved by adding a second npn transistor to act as a |
| 70 |
shorting switch across the capacitor, driven by the other serial output: |
| 71 |
|
| 72 |
|
| 73 |
. . |
| 74 |
. . ----------- +5V |
| 75 |
RTS ----------------- . | |
| 76 |
. | . Rl reset |
| 77 |
. | T1 . | |\logic |
| 78 |
. Rt | ------RWL--------| -------> |
| 79 |
. | b| /c . |/ |
| 80 |
. T2 --|---Rb---|/ . |
| 81 |
. | / | |\ . (m/b wiring typical |
| 82 |
. b| /c | | \e . only - YMMV!) |
| 83 |
DTR ------Rb--|/ C | . |
| 84 |
. |\ | | . |
| 85 |
. | \e | | . |
| 86 |
. | | | . |
| 87 |
SG ----------------------------------RWG------------- 0V |
| 88 |
. . |
| 89 |
. . stonith'ed host |
| 90 |
stonith->.<--------- RC circuit ------->.<---- RWL = reset wire live |
| 91 |
host . . RWG = reset wire ground |
| 92 |
|
| 93 |
|
| 94 |
Now when RTS goes high it can only charge up C and cause a reset if DTR is |
| 95 |
simultaneously kept low - if DTR goes high, T2 will switch on and discharge |
| 96 |
the capacitor. Only a very unusual piece of software e.g. the rcd_serial |
| 97 |
plugin, is going to achieve this (rather bizarre) combination of signals |
| 98 |
(the "meaning" of which is something along the lines of "you are clear to |
| 99 |
send but I'm not ready"!). T2 can be another BC108 and with Rb the same. |
| 100 |
|
| 101 |
RS232 signal levels are typically +-8V to +-12V so a 16V rating or greater |
| 102 |
for the capacitor is sufficient BUT NOTE that a _polarised_ electrolytic should |
| 103 |
not be used because the voltage switches around as the capacitor charges. |
| 104 |
Nitai make a range of non-polar aluminium electrolytic capacitors. A 16V 47uF |
| 105 |
radial capacitor measures 6mm diameter by 11mm long and along with the 3 |
| 106 |
resistors (1/8W are fine) and the transistors, the whole circuit can be built |
| 107 |
in the back of a DB9 serial "plug" so that all that emerges from the plug are |
| 108 |
the 2 reset wires to go to the stonith'ed host's m/b reset pins. |
| 109 |
|
| 110 |
NOTE that with these circuits the reset wires are now POLARISED and hence |
| 111 |
they are labelled RWG and RWL above. You cannot connect to the reset pins |
| 112 |
either way round as you can when connecting a manual reset switch! You'll |
| 113 |
soon enough know if you've got it the wrong way round because your machine |
| 114 |
will be in permanent reset state ;-( |
| 115 |
|
| 116 |
|
| 117 |
How to find out if your motherboard can be reset by these circuits |
| 118 |
------------------------------------------------------------------ |
| 119 |
|
| 120 |
You can either build it first and then suck it and see, or, you need a |
| 121 |
multimeter. The 0V rail of your system is available in either |
| 122 |
of the 2 black wires in the middle of a spare power connector (one of |
| 123 |
those horrible 4-way plugs which you push with difficulty into the back |
| 124 |
of hard disks, etc. Curse IBM for ever specifying such a monstrosity!). |
| 125 |
Likewise, the +5V rail is the red wire. (The yellow one is +12V, ignore |
| 126 |
this.) |
| 127 |
|
| 128 |
First, with the system powered down and the meter set to read ohms: |
| 129 |
|
| 130 |
check that one of the reset pins is connected to 0V - this then |
| 131 |
is the RWG pin; |
| 132 |
|
| 133 |
check that the other pin (RWL) has a high resistance wrt 0V |
| 134 |
(probably > 2M) and has a small resistance wrt to +5V - between |
| 135 |
0.5k and 10k (or higher, doesn't really matter) will be fine. |
| 136 |
|
| 137 |
Second, with the system powered up and the meter set to read Volts: |
| 138 |
|
| 139 |
check that RWG is indeed that i.e. there should be 0V between it |
| 140 |
and the 0V rail; |
| 141 |
|
| 142 |
check that RWL is around +5V wrt the 0V rail. |
| 143 |
|
| 144 |
If all this checks out, you are _probably_ OK. However, I've got one |
| 145 |
system which checks out fine but actually won't work. The reason is that |
| 146 |
when you short the reset pins, the actual current drain is much higher than |
| 147 |
one would expect. Why, I don't know, but there is a final test you can do |
| 148 |
to detect this kind of system. |
| 149 |
|
| 150 |
With the system powered up and the meter set to read milliamps: |
| 151 |
|
| 152 |
short the reset pins with the meter i.e. reset the system, and |
| 153 |
note how much current is actually drained when the system is in |
| 154 |
the reset state. |
| 155 |
|
| 156 |
Mostly you will find that the reset current is 1mA or less and this is |
| 157 |
fine. On the system I mention above, it is 80mA! If the current is |
| 158 |
greater than 20mA or so, you have probably had it with the simple circuits |
| 159 |
above, although reducing the base bias resistor will get you a bit further. |
| 160 |
Otherwise, you have to use an analog switch (like the 4066 - I had to use 4 |
| 161 |
of these in parallel to reset my 80mA system) which is tedious because then |
| 162 |
you need a +5V supply rail to the circuit so you can no longer just build it |
| 163 |
in the back of a serial plug. Mail me if you want the details. |
| 164 |
|
| 165 |
With the circuit built and the rcd_serial plugin compiled, you can use: |
| 166 |
|
| 167 |
stonith -t rcd_serial -p "testhost /dev/ttyS0 rts XXX" testhost |
| 168 |
|
| 169 |
to test it. XXX is the duration in millisecs so just keep increasing this |
| 170 |
until you get a reset - but wait a few secs between each attempt because |
| 171 |
the capacitor takes time to discharge. Once you've found the minimum value |
| 172 |
required to cause a reset, add say 200ms for safety and use this value |
| 173 |
henceforth. |
| 174 |
|
| 175 |
Finally, of course, all the usual disclaimers apply. If you follow my |
| 176 |
advice and destroy your system, sorry. But it's highly unlikely: serial |
| 177 |
port outputs are internally protected against short circuits, and reset pins |
| 178 |
are designed to be short circuited! The only circumstance in which I can |
| 179 |
see a possibility of damaging something by incorrect wiring would be if the |
| 180 |
2 systems concerned were not at the same earth potential. Provided both |
| 181 |
systems are plugged into the same mains system (i.e. are not miles apart |
| 182 |
and connected only by a very long reset wire ;-) this shouldn't arise. |
| 183 |
|
| 184 |
John Sutton |
| 185 |
[email protected] |
| 186 |
October 2002 |