<html><head><title>Re: [smokeping-users] time error in VM - RRDs out of sync?</title>
</head>
<body>
<span style=" font-family:'Courier New'; font-size: 9pt;">Just tinkering a little - as I'm doing some other work that just happens to be related to fping and TZ data, etc.<br>
I *think* the timestamp in the RRD is in Universal [UTC] time, not local. [I haven't looked, but guessing.]<br>
<br>
If I'm right, and it IS in UTC, make sure you've wanted until after 6.40a 7/6/2019 *UTC* time. [i.e. You'll have to calculate your own offset to make sure you're past that epoch time of 1562395151. Or, just trial-and-error, by watching the debug data till you're past it - it's in seconds since 1/1/1970.]<br>
<br>
-Greg<br>
<br>
<br>
</span><table>
<tr>
<td width=2 bgcolor= #0000ff><br>
</td>
<td><span style=" font-family:'courier new'; font-size: 9pt;">Just guessing here - though a somewhat educated guess.<br>
It looks like you updated the RRD when the system clock read ~6.40a 7/6/19 - and the clock was wrong at that point.<br>
<br>
But, now that you've "fixed" the clock's time it's now a date/time that's *earlier* than the last update in the RRD.<br>
<br>
I think if you simply wait to restart smokeping until after ~6.40a on 7/6/2019, you'll now have a clock date that is beyond the last RRD update and it will accept it. [The updates/new records.]<br>
<br>
IMO, that's the most straight-forward way to "fix" your problem.<br>
<br>
I suppose, alternatively, you could remove the RRD's that have the wrong timestamps, or alter them somehow - but I'd guess that is fraught with problems, and is a LOT more work. Simply waiting ~24h seems a lot less work. :) With the obvious downsides that you won't get updates or accurate records in the interim, and the old/current records during when the clock was wrong will be confusing - if you ever need to really dig into that data.<br>
<br>
-Greg<br>
<br>
<br>
<br>
</span><table>
<tr>
<td width=2 bgcolor= #0000ff><br>
</td>
<td><span style=" font-family:'calibri'; font-size: 11pt;">I moved my smokeping hosting system to a new VM, and now it is failing with an error from what seems to be a difference in time/clocks in VMs, what seems to be a known problem.<br>
<br>
I think I reset the clocks properly, and the RTC is now in sync, but perhaps the RRD history files are now out of sync?<br>
<br>
How to recover??<br>
<br>
Calling RRDs::update(/var/lib/smokeping/World/MUM_CS/SomeSystem.rrd --template uptime:loss:median:ping1:ping2:ping3:ping4:ping5:ping6:ping7:ping8:ping9:ping10:ping11:ping12:ping13:ping14:ping15:ping16:ping17:ping18:ping19:ping20 1562366484:U:0:1.0400000000e-03:5.8000000000e-04:6.0000000000e-04:6.1000000000e-04:6.1000000000e-04:6.2000000000e-04:6.4000000000e-04:7.1000000000e-04:7.6000000000e-04:7.8000000000e-04:9.1000000000e-04:1.0400000000e-03:1.0500000000e-03:1.1000000000e-03:1.8100000000e-03:3.1200000000e-03:5.9300000000e-03:6.4400000000e-03:7.9900000000e-03:8.0500000000e-03:1.0810000000e-02)<br>
RRDs::update ERROR: /var/lib/smokeping/World/MUM_CS/ SomeSystem.rrd: illegal attempt to update using time 1562366484 when last update time is 1562395151 (minimum one second step)<br>
<br>
Do I need to delete all current RRD files? (and lose all history!)<br>
<br>
root@alpha:/etc/smokeping/config.d# timedatectl<br>
Local time: Fri 2019-07-05 17:44:27 CDT<br>
Universal time: Fri 2019-07-05 22:44:27 UTC<br>
RTC time: Fri 2019-07-05 22:45:08<br>
Time zone: America/Chicago (CDT, -0500)<br>
Network time on: yes<br>
NTP synchronized: yes<br>
RTC in local TZ: no<br>
</td>
</tr>
</table>
</td>
</tr>
</table>
<br><br>
<span style=" font-family:'arial'; color: #c0c0c0;"><i>-- <br>
Gregory Sloop, Principal: Sloop Network & Computer Consulting<br>
Voice: 503.251.0452 x82<br>
EMail: </i></span><a style=" font-family:'arial';" href="mailto:gregs@sloop.net">gregs@sloop.net</a><br>
<a style=" font-family:'arial';" href="http://www.sloop.net">http://www.sloop.net</a><br>
<span style=" font-family:'arial'; color: #c0c0c0;"><i>---</body></html>