We have all the sensors and the data system running in the staging area. We've added one Licor 7500 to what was last on the tower, for a total of 4 Licors. The data cables, and Licor power cables are different however, since the originals were left on the tower.
After setting it up, I increased the 4 Licor 7500s from 10 samples/sec at 9600 baud to 20 samples/sec at 19200 baud.
The spurious interrupts are happening, sometimes spiking to above 100:
Jun 28 21:27:02 manitou kernel: viper_irq_handler: irq=2 spurious= 326001, #/sec=102 Jun 28 21:27:21 manitou kernel: viper_irq_handler: irq=2 spurious= 328001, #/sec=109 Jun 28 21:29:06 manitou kernel: viper_irq_handler: irq=2 spurious= 338001, #/sec=102 Jun 28 21:29:46 manitou kernel: viper_irq_handler: irq=2 spurious= 342001, #/sec=106 Jun 28 21:29:56 manitou kernel: viper_irq_handler: irq=2 spurious= 343001, #/sec=102 Jun 28 21:30:38 manitou kernel: viper_irq_handler: irq=2 spurious= 347001, #/sec=103 Jun 28 21:30:48 manitou kernel: viper_irq_handler: irq=2 spurious= 348001, #/sec=101 Jun 28 21:31:28 manitou kernel: viper_irq_handler: irq=2 spurious= 352001, #/sec=111 Jun 28 21:31:58 manitou kernel: viper_irq_handler: irq=2 spurious= 355001, #/sec=104 root@manitou root# uptime 21:44:36 up 1:43, 1 user, load average: 0.16, 0.11, 0.09
The kernel issues a "spurious" message when they occur more often than 100/sec. The system came up at 21:44 - 1:43 = 20:01. At the time of the last message above, the system had been up for 1 hour 31 minutes. 355001 interrupts in 1 hour 31 minutes averages to 65/sec.
The above test is with the same kernel (2.6.16) and CPU (viper 4) as was deployed on the tower.