Aberrant time display in Monitoring graph

provels

@johnpoz
Yes, time is same. PFSense is VM and uses its NTP client. It does not use host time sync. Host uses it's own NTP client (Meinberg, not Windows). All match +/- ms.

Thanks for the assist.

johnpoz

This is the graph your looking at right

If you zoom into say 1 hour

provels

@johnpoz
Yes, here's the 1 hour, wacky:

johnpoz

So your in the future ;) Maybe it can tell the future and knows your going to have an outage on the 14th ;)

edit:
So clearly pfsense thinks its the 14th..

From cmd line check..

[21.05-RELEASE][admin@sg4860.local.lan]/root: date
Fri Jun 11 06:51:54 CDT 2021
[21.05-RELEASE][admin@sg4860.local.lan]/root:

What does say the firewall log show for date and time?

provels

@johnpoz Screw it, time to whack the RRD data I think.

Graph locked at 6/14 0629...

[2.5.1-RELEASE][root@fw.workgroup]/root: date
Fri Jun 11 07:09:23 CDT 2021

johnpoz

Well odd for sure - currently drawing a blank.. Have you tried resetting the data?

Do all the other monitor graphs show the same in the future date?

provels

@johnpoz said in Aberrant time display in Monitoring graph:

Do all the other monitor graphs show the same in the future date?

Yes they did, but cleared data now. Thanks for your attention anyway. Over and out.

johnpoz

So the reset data fixed it?

Curious how it got mixed up.. I would have to think at some point pfsense had wrong date..

provels

@johnpoz
Yep. go figure. fortunately, I'm only responsible to myself for SLA...

johnpoz

Very strange, I wonder if @jimp might be able to shed some light on how it could happen?

Or maybe someone else that is more versed in the code on how that data is stored and retrieved can shed some light on how it happened.

provels

@johnpoz
From what I've read, RRD Summary and Monitoring both use the same data, so I'd chalk it up to crud and corruption in the DB. Perhaps it is something to do with the virtual environment and host reboots. Normally, I shut pfSense down manually before a reboot, but I've also let the host save state recently. Back to shutting down FW first.

jimp

Hard to say what might have happened there. Perhaps the VM clock jumped ahead at some point, had data added to the RRD file, and then the clock was set back.

Last I knew, RRD did not take kindly to future dates in the RRD files and it may have caused it to stop adding data since nothing was newer than the latest timestamp.

johnpoz

@jimp said in Aberrant time display in Monitoring graph:

caused it to stop adding data since nothing was newer than the latest timestamp.

So if he would of just left it, once the actual date got beyond the 14th he was seeing - it could starting added new data?

jimp

I suspect it would have. That's what has happened in the past during similar situations when I've tried it. Granted the last time I tried any time-based shenanigans with RRD was many many years ago. I don't think it's changed that much in its core though.