View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0002697||SpeedFan||Logging and charting||public||2018-07-25 17:34||2018-12-26 16:00|
|Platform||Desktop PC||OS||Windows 7 64-bit||OS Version||6.1.7601|
|Summary||0002697: Temperature charts are invalid when monitoring SSDs|
|Description||I have 2 Samsung SSDs in my system, one is M.2, the other is SATA.|
In the chart page, if I monitor temperatures *only* on MB and GPU, the charts display normally. However, if I enable either of the SSD monitors, the graph is corrupted as shown below.
|Steps To Reproduce||> start Speedfan|
> on chart page, select MB and GPU temperatures
> let Speedfan run long enough to fill chart
> observe that chart displays all temperature graphs normally
> select one of HD1/HD2
> observe that the chart is corrupted, as shown in the attached image
|Tags||No tags attached.|
|Motherboard Model||GIGABYTE X79-UP4|
|Video Card Model||GeForce GTX 980|
The graph is not "corrupt", as such.
The issue lays with the different sampling rates of the variables.
The graph is displaying variables with different measurement intervals and lengths of plotted history, so they look ... messy ... on the same timescale.
If you look at the Start time & End time values below the graph you'll see that the "Start time" is different for various sensor types.
The CPU/GPU/Motherboard etc sensors are being read and plotted every 3 seconds, the disk SMART values every minute (which is reasonable - SMART was designed as a long-term/low-overhead monitoring protocol, so up to the second polling is overkill).
Each sensor has the same number (250) of data-points stored so the available 13 minutes of CPU/GPU data is all up at the right side of the timeline and has no data to backfill the longer 5.5 hour timescale of the SMART data.
Okay, I understand this explanation, and what is occurring.
Nevertheless, though, the basic issue remains; the graph isn't usable if I'm plotting all devices.
Is there any reason that the disk data couldn't be *plotted* at the same frequency as the other devices, even though it isn't *sampled* as often?? That would still cover 13 1-minute updates in the 13-minute graph, but it would make the overall graph usable.
It would be nice, I guess, but it's been an issue since ... forever ... and I've just grown used to and accepted "That's Just The Way It Is[tm]" and Alfredo doesn't seem like he's going to be doing anything with the Speedfan project any time soon.
As a sorta-kinda workaround you could use the logging facility under the Configuration->Log tab and dump the data out to an ongoing logfile for a more capable external graphing solution to deal with.
Whether that's worthwhile ... Your Mileage May Vary :)
Ah, I just realized, you aren't a developer, you're another user!!
I really appreciate your insight into what is going on with this issue, because I'd really been stumped by it; your explanation make it clear...
It's a shame Alfredo didn't release the source code once he moved on from it, so someone else could pick it up. I tried looking for some other replacement program, but nothing else has the graphing function, which is quite valuable !!
I'll probably stick with things as they are, and try to find some other monitor for the SSDs; in any case, they are less important (for real-time monitoring) than the CPUs and GPU.
|2018-07-25 17:34||Gorlash||New Issue|
|2018-07-25 17:34||Gorlash||Status||new => assigned|
|2018-07-25 17:34||Gorlash||Assigned To||=> alfredo|
|2018-07-25 17:34||Gorlash||File Added: Speedfan.with.SSDs.jpg|
|2018-08-08 09:05||kurai||Note Added: 0008609|
|2018-08-08 09:06||kurai||Note Edited: 0008609|
|2018-08-08 15:50||Gorlash||Note Added: 0008610|
|2018-08-08 16:02||kurai||Note Added: 0008611|
|2018-08-08 16:32||Gorlash||Note Added: 0008612|