![Code*
#!/usr/bin/python
import dhtreader, time, os, sys, os.path
type=22
pins =(18,)
dhtreader.init()
time.sleep(3)
for loop in range(SAMPLECOUNT):
for pin in pins:
time.sleep(0.5)
val = dhtreader.read(type,pin)
if val:
temp, hum = val
sensors[pin]["count"] += 1
sensors[pin]["temperature"] += temp
sensors[pin]["humidity"] += hum
time.sleep(0.5)
for pin in pins:
if sensors[pin]["count"] > 0:
sensors[pin]["tempoutput"]=str(sensors[pin]["temperature"] /
sensors[pin]["count"])
sensors[pin]["humoutput"]=str(sensors[pin]["humidity"] /
sensors[pin]["count"])
else:
sensors[pin]["tempoutput"]="ERR"
sensors[pin]["humoutput"]="ERR"
sensors[pin]["errors"]=str(SAMPLECOUNT-
sensors[pin]["count"])
# output averages
fpw.write( "%s,"%time.ctime() +
','.join(["%s,%s"%(sensors[x]["tempoutput"],sensors[x]["humoutput"
]) for x in pins])+"n")
sys.stderr.write("%s, samples=%d,
"%(time.ctime(),SAMPLECOUNT)+','.join(["S%d_errorcount=%s"%
(x,sensors[x]["errors"]) for x in pins])+"n”)
*Complete source available in comments
on this slide.](https://image.slidesharecdn.com/connect2015-smokingsalmon-150726214123-lva1-app6892/85/Connect-2015-Smoking-Salmon-with-Splunk-and-Raspberry-Pi-6-320.jpg)
Connect 2015 Smoking Salmon with Splunk and Raspberry Pi
- 1. Smoking Salmon with Splunk and Raspberry Pi Greg Dostatni
- 2. Assumptions Cold Smoking Requires temperatures between 20°C and 30°C Raspberry Pi Tiny computer used by hobbyists for just about anything Splunk Software used for searching, visualizing and analyzing machine generated data
- 3. Equipment
- 4. Electronics
- 6. Code* #!/usr/bin/python import dhtreader, time, os, sys, os.path type=22 pins =(18,) dhtreader.init() time.sleep(3) for loop in range(SAMPLECOUNT): for pin in pins: time.sleep(0.5) val = dhtreader.read(type,pin) if val: temp, hum = val sensors[pin]["count"] += 1 sensors[pin]["temperature"] += temp sensors[pin]["humidity"] += hum time.sleep(0.5) for pin in pins: if sensors[pin]["count"] > 0: sensors[pin]["tempoutput"]=str(sensors[pin]["temperature"] / sensors[pin]["count"]) sensors[pin]["humoutput"]=str(sensors[pin]["humidity"] / sensors[pin]["count"]) else: sensors[pin]["tempoutput"]="ERR" sensors[pin]["humoutput"]="ERR" sensors[pin]["errors"]=str(SAMPLECOUNT- sensors[pin]["count"]) # output averages fpw.write( "%s,"%time.ctime() + ','.join(["%s,%s"%(sensors[x]["tempoutput"],sensors[x]["humoutput" ]) for x in pins])+"n") sys.stderr.write("%s, samples=%d, "%(time.ctime(),SAMPLECOUNT)+','.join(["S%d_errorcount=%s"% (x,sensors[x]["errors"]) for x in pins])+"n”) *Complete source available in comments on this slide.
- 7. Dry Cure 4 cups brown sugar 1 cup salt Spices (optional) Video instructions at http://youtu.be/Jk3xl6H9o94
- 10. Motivation Salmon economics Salmon Filet $9 / lbs. Typical batch 4 lbs. Finished weight 2.8 lbs. C. Smoked Salmon $ 40 - 45 / lbs. Savings $ 70 - 85 CFO’s approval … Priceless
- 11. Future Plans
- 12. Future Plans #2
Editor's Notes
- #4: Inexpensive diy smoker
- #6: Diy breakout board for rapidly connecting sensors.
- #7: #!/usr/bin/python import dhtreader import time import os, sys, os.path type=22 #pins=(23,24,25) pins =(18,) LOGDIR="/var/log/sensor" LOGFNAME = "sensors.csv" SAMPLECOUNT = 5 HEADER="Timestamp," + ','.join(["S%d_Temp,S%d_Humidity"%(x,x) for x in pins])+"\n" dhtreader.init() time.sleep(3) if not os.path.exists(os.path.join(LOGDIR, LOGFNAME)) : fpw=open(os.path.join( LOGDIR ,LOGFNAME),"a+") fpw.write(HEADER) else: fpw=open(os.path.join( LOGDIR ,LOGFNAME),"a+") # Sampling loop. # Sample 5 times each sensor , 1 second between sample runs # Record averages in CSV, Output error rate to stderr sensors={} for pin in pins: sensors[pin]={"count":0,"temperature":0,"humidity":0} for loop in range(SAMPLECOUNT): for pin in pins: time.sleep(0.5) val = dhtreader.read(type,pin) if val: temp, hum = val sensors[pin]["count"] += 1 sensors[pin]["temperature"] += temp sensors[pin]["humidity"] += hum time.sleep(0.5) for pin in pins: if sensors[pin]["count"] > 0: sensors[pin]["tempoutput"]=str(sensors[pin]["temperature"] / sensors[pin]["count"]) sensors[pin]["humoutput"]=str(sensors[pin]["humidity"] / sensors[pin]["count"]) else: sensors[pin]["tempoutput"]="ERR" sensors[pin]["humoutput"]="ERR" sensors[pin]["errors"]=str(SAMPLECOUNT-sensors[pin]["count"]) # output averages fpw.write( "%s,"%time.ctime() + ','.join(["%s,%s"%(sensors[x]["tempoutput"],sensors[x]["humoutput"]) for x in pins])+"\n") sys.stderr.write("%s, samples=%d, "%(time.ctime(),SAMPLECOUNT)+','.join(["S%d_errorcount=%s"%(x,sensors[x]["errors"]) for x in pins])+"\n" )
- #10: Spike corresponds to when I accidently dumped a lot of snow on one of the sensors.
- #12: Without data you don’t know what you get. Above: Temperature and humidity of my dryer (two loads). There is a fair amount of time spent just warming up the clothes without removing any more water. Below: A small bar fridge we use to store breast milk for the little one. The temperature fluctuates over a 5 degree range, which is much higher than expected. I realize that the bottle of milk will have a much better stability, but the range did surprise me.
- #13: This is a 3d model of my house. One of the perksof being married to a structural engineer. I want to make it into a house dashboard inside of our home Splunk.