SEQMIN1 Explanation

[dgunnell]

What is the format for these sequence entries? For example, there are numerous entries in the SEQMIN1_* data for each of the variables so what is this representing? Is this really more that a one minute sample? Is there detailed documentation explaining all the SEQ data fields somewhere?

Thanks,

Dave

[admin]

SEQ things have been introduced to cut graph generating data for wdlive. It is a sequence of data, each item separated by "_".

time stamps for that can be taken from corresponding "localdate" or "utcdate" line.

[tomo]

for which period of time representing data seqmin1_*? 60 items of data for one hour? from time to time i have between 55 and 60 items in one hour

[cgn]

The seqmin1_ localdate will be updated every minute.
But I noticed that the values in seqmin1_wind0... _rain0 etc. will be updated only every 3 minutes.
So the first value in the sequence can be a value from 2 minutes ago or from 3 or 4 minutes ago.
Example:
seqmin1_localdate 20120318134009_
seqmin1_wind0_gustspeed_kn 2.8_1.9_2.1_3.1_2.3_3.0_ ...
seqmin1_localdate 20120318134109_
seqmin1_wind0_gustspeed_kn 2.8_1.9_2.1_3.1_2.3_3.0_ ...
seqmin1_localdate 20120318134209_
seqmin1_wind0_gustspeed_kn 2.8_1.9_2.1_3.1_2.3_3.0_ ...
seqmin1_localdate 20120318134309_
seqmin1_wind0_gustspeed_kn 5.4_4.6_2.4_2.8_1.9_2.1_ ...

Is this a bug or isn't it possible to update the sequence every minute?

Another issue: Why is the windspeed given in kn and not in kmh as returned in actual_system_wind_unit ?

Meteohub version is 4.9d (Build 969).

[idociko]

I think wind is in Knots because WDlive uses knots. But you can display Kmh on your site. For example to display Kmh windspeed from 6 minutes ago past in your template: {math*[seqmin1_wind0_speed_kn@6] 1.853}. Conversion Knot>Kmh = Knots*1.853.

[cgn]

OK, thank you, this is not really a problem.
But from which time is the 1st value in the sequence of seqmin1_wind0_speed_kn ???
It is not the time given in seqmin1_localdate, but the time two, three or four minutes ago, it's different.

[idociko]

Your right! seqmin1 updates just every 3 minutes instead every minute. I didn't use seqmin1, only seqmin5 and above, so I never noticed this failure.

This table shows what seq??? does:

seqmin1 Last 60 mins in steps of 1 minute (now -1min .. now -60 minutes)
Updated every minute by both histeval0 (might be a bug) and histeval1 - 7

seqmin15 Last 23.5 hours in steps of 15 minutes hour(now -15mins .. now -23.5 hours)
Updated every 5 minutes by histeval1 - 6

seqhour1 Last 25 hours in steps of 1 hour(now -1hr .. now -25hrs)
Updated every 5 minutes by histeval1 - 8

seqhour6 6 hour interval, total 23 steps (now .. now -~138 hrs)
Updated every 6 hours (00:13/06:13/12:13/18:13) by histeval2 - 3

seqday1 Last 15 days in steps of 1 day(yesterday .. today -23 days)
Updated every 6 hours (00:13/06:13/12:13/18:13) by histeval2 - 2

seqmonth1 Last 14 months in steps of 1 month(last month .. today -14 months)
Updated every day (04:47) by histeval3 - 2

So, if you want the data from wind0 2 minutes ago use: [seqmin1_wind0_speed_kn@2] (second item in array).

I hope it's clear to you, if not recall.

Jos

[cgn]

Your right! seqmin1 updates just every 3 minutes instead every minute.

and

Updated every minute by both histeval0 (might be a bug) and histeval1 - 7

don't fit together.

So, if you want the data from wind0 2 minutes ago use: [seqmin1_wind0_speed_kn@2] (second item in array).

When data are updated only every 3 minutes, the second item in the array can be from 2 or 3 or 4 minutes ago!!!

What is "histeval0" or "histeval1-7"?

[idociko]

Instead of

Updated every minute

I should have wrote:

Should updated every minute

HistevalX are internal processes that updates the values.

[cgn]

Thank you for your answer.

I hope Boris can give an explanation for the "3-minutes-updates".

[YJB]

If you are running a NSLU2, this is your answer:

Code: Select all

#!/bin/sh
if [ "$METEOHUBHW" == "NSLU2" ]
then
  mod=$(((`date +"%s"` / 60) % 3))
  if [ $mod -gt 0 ]
  then
    exit 0
  fi
fi

Or in other words, to reduce the load on the, relative slow, NSLU2, histeveval0 is only running every 3 minutes

[idociko]

I'm running Alix1D.

[cgn]

I have running Meteohub on a sheevaplug, system load is not a problem.

And I'm a "Windows-only" -user, so I don't know how to use Unix/Linux-Scripts as shown in your code.

[cgn]

Is it possible to set this histeval (for seqmin1) to updating every minute on a sheevaplug?

[YJB]

Hi,

As shown in the script, it should run every minute, unless your system is a NSLU2.

I'm not sure hwo to troubleshoot, since you've indicated that you are a Windows only user.

What I normally would do (and I verified this on my own MH) is:
- Check what MH think t is running on:
echo $METEOHUBHW
On an Alix:

Code: Select all

X86

And if it is reporting NSLU2, then your MH is confused.

If it is reporting something else, my next step would be to verify that the script is really running every minute, bu logging a simple timestamp (in the histeval0
after the METEOHUBHW test):
date >> /tmp/histeval0.log

And check after a couple of minutes the output.

Code: Select all

Wed Mar 21 10:46:01 CET 2012
Wed Mar 21 10:47:01 CET 2012
Wed Mar 21 10:48:01 CET 2012
Wed Mar 21 10:49:02 CET 2012
Wed Mar 21 10:50:02 CET 2012
Wed Mar 21 10:51:01 CET 2012
Wed Mar 21 10:52:02 CET 2012

But, you will need to know a bit LInux or find somebody who does. Also, make sure that you revert the change to the script afterwards, otherwise you will be running out of space eventually.