Weewx no longer updating...

[Alec Bennett]

My weather station is still building reports, but it's stuck on the date June 20 (5 weeks ago). However the data in the real time gauge data is correct.

I don't see anything interesting in /var/log/syslog.

Any advice on how to troubleshoot this?

[Alec Bennett]

Here's an interesting one, putting this in a skin file outputs "2:45pm on Tuesday June 20, 2017", even though today is June 27:

$current.dateTime.format("%-I:%M%p").lower() on $current.dateTime.format("%A %-B %-d, %Y")

(And oops, I meant to say June 20 is 1 week ago, not 5 weeks ago. So for some it's stuck on the date 1 week ago.).

--
You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/P4Rc4V4_dRU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[gjr80]

Hi,
 

Any advice on how to troubleshoot this?

Yes, set debug=1, stop/start weeWX and post the log from startup until a couple of archive periods have completed. You would be surprised what you can garner about a system from the log.

Gary

[Thomas Keffer]

You're giving us almost nothing to go on. No log, no weather station type, etc.

On the off chance you have a Vantage, did you see the section WeeWX generates HTML pages, but it does not update them?

-tk

On Tue, Jun 27, 2017 at 4:16 PM, Alec Bennett <wryb...@gmail.com> wrote:

--
You received this message because you are subscribed to the Google Groups "weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+unsubscribe@googlegroups.com.

[Alec Bennett]

Oops sorry about that, should have mentioned, it's a Davis Vantage Pro2.

Thanks for the troubleshooting tips. I rebooted the console since "wee_device --clear-memory"  wasn't getting any response. Now I'm getting the following errors, hopefully it just needs to run for awhile?

Jun 27 17:05:36 weather weewx[2079]: engine: Starting up weewx version 3.6.2

Jun 27 17:05:36 weather weewx[2079]: engine: Clock error is -0.05 seconds (positive is fast)

Jun 27 17:05:36 weather weewx[2079]: engine: Starting main packet loop.

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: **** Traceback (most recent call last):

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 882, in process_packet

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: ****     data = self.calculate(cached_packet)

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 1466, in calculate

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: ****     fromBearing = max((self.windDirAvg-d) if ((d-self.windDirAvg) < 0 and s > 0) else None for x,y,s,d,t in self.buffer.wind_dir_list)

Jun 27 17:05:36 weather weewx[2079]: rtgdthread: **** ValueError: max() arg is an empty sequence

Jun 27 17:05:40 weather weewx[2079]: vantage: LOOP try #1; error: Expected to read 99 chars; got 0 instead

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: **** Traceback (most recent call last):

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 882, in process_packet

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: ****     data = self.calculate(cached_packet)

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 1466, in calculate

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: ****     fromBearing = max((self.windDirAvg-d) if ((d-self.windDirAvg) < 0 and s > 0) else None for x,y,s,d,t in self.buffer.wind_dir_list)

Jun 27 17:05:40 weather weewx[2079]: rtgdthread: **** ValueError: max() arg is an empty sequence

Jun 27 17:05:44 weather weewx[2079]: vantage: LOOP try #2; error: Expected to read 99 chars; got 0 instead

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: **** Traceback (most recent call last):

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 882, in process_packet

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: ****     data = self.calculate(cached_packet)

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 1466, in calculate

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: ****     fromBearing = max((self.windDirAvg-d) if ((d-self.windDirAvg) < 0 and s > 0) else None for x,y,s,d,t in self.buffer.wind_dir_list)

Jun 27 17:05:44 weather weewx[2079]: rtgdthread: **** ValueError: max() arg is an empty sequence

Jun 27 17:05:48 weather weewx[2079]: vantage: LOOP try #3; error: Expected to read 99 chars; got 0 instead

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: **** Traceback (most recent call last):

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 882, in process_packet

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: ****     data = self.calculate(cached_packet)

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: ****   File "/home/weewx/bin/user/rtgd.py", line 1466, in calculate

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: ****     fromBearing = max((self.windDirAvg-d) if ((d-self.windDirAvg) < 0 and s > 0) else None for x,y,s,d,t in self.buffer.wind_dir_list)

Jun 27 17:05:48 weather weewx[2079]: rtgdthread: **** ValueError: max() arg is an empty sequence

Jun 27 17:05:52 weather weewx[2079]: vantage: LOOP try #4; error: Expected to read 99 chars; got 0 instead

Jun 27 17:05:52 weather weewx[2079]: vantage: LOOP max tries (4) exceeded.

Jun 27 17:05:52 weather weewx[2079]: engine: Caught WeeWxIOError: Max tries exceeded while getting LOOP data.

Jun 27 17:05:52 weather weewx[2079]:     ****  Waiting 60 seconds then retrying...

--
You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/P4Rc4V4_dRU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+unsubscribe@googlegroups.com.

[gjr80]

The rtgd error is known (issue 6) and was fixed in v0.2.13 released 7 May. I believe the issue will go away after the first archive record is saved after startup. Clearly there are some other issues in the log, you might want to temporarily disable rtgd by commenting out user.rtgd.RealtimeGaugeData in the report_services setting under [Engine] [[Services]] in weewx.conf - one less issue to muddy the waters. If you are getting no data from the station then no reports will be run and the old ones from 20 June will remain. Once you do get things sorted I suggest you upgrade rtgd to v0.2.13.

Gary

To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/P4Rc4V4_dRU/unsubscribe.

To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[Alec Bennett]

Thanks for the phenomenal help, working perfectly again. For anyone else who comes down this path, I rebooted (pulled batteries) the Vantage Pro2 console and ran "wee_device --clear-memory". Not sure if that second step was necessary.

What's odd is that the interface on the console itself was working well (displaying the correct data) even though the USB interface was apparently dead for a week. And if weewx was throwing an error, I didn't see it. The only symptom of the error I could find was that all the data was old.

To detect this in the future I'm thinking about making a cronjob that checks the connection once a day and fires off an email if it's broken. Since wee_device returned an error when the connection was down, I'm thinking just a call to wee_device with no params would work to detect the error state. Or maybe there's a better way? Thoughts?

 

To unsubscribe from this group and all its topics, send an email to weewx-user+unsubscribe@googlegroups.com.

[Alec Bennett]

My Vantage Pro2 is stuck uploading week old data again. What's odd is that the "real time gauge" (RTG) data is working just fine, so communication between weewx and the Davis console receiver appears to be working. 

Unfortunately I can't be onsite to fix it for a couple of days, but any idea what might be causing this issue? Some details:

- Davis Vantage Pro2

- data is frozen at a week old

- the fact that the rtg data is still working makes me suspect it wasn't the reboot?

- this is the second time this happened

- last time I fixed it by rebooting the Davis console receiver, and running "wee_device --clear-memory". I'm not sure which step fixed things but everything worked again after that 

I suppose I could make a script that checks the age of the data as seen by the generation time and then runs wee_device --clear-memory? But of course I'd rather figure out what's causing the issue in the first place. 

To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[Andrew Milner]

what does the log say???

[gjr80]

Hi Alec,

Quite possibly the same issue again. When the console suffers from corrupted memory it is usually just archive records from the console that are affected; loop data is usually unaffected (not sure if this is always the case though). If you do indeed have corrupted console memory then this would explain why you continue to see the RTG data updating. The RTG service uses loop data to generate its output and is thus unaffected by the corrupted memory. However, the traditional html pages are generated during the report cycle based upon archive records from the console (assuming you have record_generation = hardware in weewx.conf), your system is not receiving any archive records so data (other than date-time) in the generated reports doesn't change.

Having some sort of cron job to clear the memory is a bit of a kludge, worthwhile trying to sort out the issue, consoles should not suffer from corrupted memory on a regular basis unless something is wrong.

Gary

[Alec Bennett]

Any idea what would corrupt the console memory?

And am I just unlucky that this has now happened twice in a month, or is this a common issue with the Davis Vantage Pro2?

[Thomas Keffer]

Twice in one month is very unusual. My logger is about 5 years old and it's only happened once. 

Corrupted memory is usually, but not always, correlated with messing with the internal clock, such as by changing time zones or turning DST on and off.

You can give Davis a call and see what they think.

Alternatively, you can switch to software record generation. Then the archive records will be calculated by weewx, rather than downloaded from the logger.

-tk

To unsubscribe from this group and all its topics, send an email to weewx-user+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "weewx-user" group.

To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+unsubscribe@googlegroups.com.

[gjr80]

Power can cause problems. I have had my system for 5 odd years and only once or twice max has it happened to me and each time it has coincided with something happening power wise. Power is pretty stable here but we do get the ocassional storm related outage. We have had a few scheduled outages and I just turn off the mains beforehand and run off batteries rather than risk it it. I do have a cabled system (rather than wireless), not sure that has any bearing though.

Gary

[John S]

I know this is an old thread, but it helped me diagnose why my weewx 4.2.0 web site was repeating the data from a set date and not advancing.

I had brought up an old Raspberry Pi with weewx configured to run at start up. I forgot that it was running weewx and was merrily configuring it for another task.

I wondered why my main weewx instance was losing chunks of data, literally, 30 minute time chunks not showing up on the website.  I didn't see anything weird in the main instance logs.

After shutting down the PI,  the main weewx instance (on ubuntu) would not advance past the PI's shutdown time. I had to clear the data on the console with wee-device --clear-memory which reset everything and the data flows again. (Use caution - it does reset all data.)

Now, to recover the PI, I guess I will shut down the Ubuntu instance, bring up the PI, remove the script from /etc/init.d,, then  stop weewx,  reboot the PI to make sure it isn't starting weewx again, then, when assured there isn't another instance running, bring the Ubuntu instance back on line.

Cheers,

John