[BUG]: Bad print quality

[theresetmaster]

HI. After installing the plugin I start seeing blobs in the print - mostly on curves. Looking closer on the print, it is noticeable that the printer have several short stops that looks like the printer is not getting information from Octoprint. It is during these stops that the blobs are forming.

When I disable the TPLinkSmartplug plugin the problem is gone. It seems that the plugin is causing the entire system to hang for a short moment.

My setup is:

• Raspberry PI 2
• Ethernet connection
• Octoprint: 1.4.0
• OctoPi 0.17.0

(I have several printers and the problem is also on other versions of OctoPi)

plugin_tplinksmartplug_debug (1).log

[jneilliii]

This may very well be related to the print progress status checking of the plug since it appears you have an energy monitoring enabled device. The older generation Pi you're using could also contribute to the issue because I haven't personally seen this type of problem using a Pi 3B or Pi 4. I'll look into making the status polling an option that can be disabled, but at the cost of not polling the energy data while printing.

[theresetmaster]

Hi. Sounds very much in line with my thought about this problem.

Making it an option to pull data would be great. Maybe also an option to set how often the data is pulled from plug ?

[jneilliii]

I made some tweaks in my latest development version that I would appreciate some testing on to see if it resolves the issue. You can install using the URL below in plugin manager.

https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.21.zip

[jneilliii]

So after 8 days I haven't heard back? I re-reviewed your posted log and it seems the print progress on these prints are happening very quickly, so I'm assuming these are small.short prints. In some cases the status check of the plug are separated by milliseconds, so I could see where this definitely could be problematic. Does the same issue occur for you with slower/larger prints that take longer than 20 minutes to complete?

[theresetmaster]

Hi

I'm sorry that I did not get back to you, but I'm having some troubles with OctoPrint at the moment that needs to be fixed first. It seems that Seems that the printers I have running Repetier firmware is having some problems with transferring data from the Raspberry PI in a steady stream, and the funny thing is that I do not see the same problem on my printer that runs mendel. The same g-code file simply prints different in quality. 

I know this does not have to do directly with you, however, just to explain why I have not been giving too much feedback on your update. 

I did however install it, but did not see any difference.

On your question on the duration of the prints, you may see prints that are stopped, but normally what I print is prints that take 2-3 hours, while some initial testing prints might be 5-15 minutes.

Best Jørgen Larsen På 12-05-2020 01:44:16, jneilliii notifications@github.com skrev: So after 8 days I haven't heard back? I re-reviewed your posted log and it seems the print progress on these prints are happening very quickly, so I'm assuming these are small.short prints. In some cases the status check of the plug are separated by milliseconds, so I could see where this definitely could be problematic. Does the same issue occur for you with slower/larger prints that take longer than 20 minutes to complete? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub [https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/issues/164#issuecomment-627023827], or unsubscribe [https://github.com/notifications/unsubscribe-auth/AC7GCRY4O3IYL4XZQMBR7NTRRCEU7ANCNFSM4MNZ3QXQ].

-- Denne e-mail blev kontrolleret for virusser af Avast antivirussoftware. https://www.avast.com/antivirus

[ybjlus]

@jneilliii I also have the same exact problem with a raspberry pi 2B. I did a clean octopi install and no problems what so ever. I then downloaded your 0.9.21 version and the stuttering started again...

[jneilliii]

Thanks @ybjlus I assume you also have an energy monitoring plug? Having that information will help isolate where this might be happening and come to a solution quicker. Unfortunately, I've been busy with some other plugin updates and haven't had a chance to really dig into this problem.

[ybjlus]

Thanks @ybjlus I assume you also have an energy monitoring plug? Having that information will help isolate where this might be happening and come to a solution quicker. Unfortunately, I've been busy with some other plugin updates and haven't had a chance to really dig into this problem.

Hi I have a tp-link hs100 standard version. But this plug is always connected to the printer also when I tested it without the plugin installed

[cbarth3]

My setup is:

Raspberry PI 3 B
Wifi
Octoprint: 1.3.12
OctoPi 0.17.0 Fresh install
Plugin v0.9.20
HS100 plug and I am seeing stops while making circles. I printed a cylindrical piece and it had lots of bumps. When I print a calibration file that has small circles at the leveling knobs I can see it stop. I disabled the plugin and printed from the SD card and do not have this problem.

I tried the v0.9.21 that you posted earlier and saw the same results. I turned on logging for this one, but not sure where the file is stored so let me know and I can post it.

[jneilliii]

@cbarth3 printing from SD probably resolved your problem, because it sounds like your issue is relative to the firmware buffer getting full. There's a large discussion about this in the community forum. If you disable the plugin and print from OctoPrint (not SD printing) the same stuttering may also happen, let me know.

[cbarth3]

@jneilliii sorry I should have been more descriptive on this. If the plug-in is disabled I do not see this issue. I have disabled the plug-in and started using the wemo version and do not see this issue with that one.

The only time the issue occurs is when the tplink plug-in is enabled.

Edited to add a picture for reference.

[jneilliii]

Ok, thanks. Your description before indicated printing from SD, and that threw me off.

[jneilliii]

@cbarth3 I just made some changes to that 0.9.21 version. If you don't mind risking bad print quality I would appreciate it if you could try with this updated version. Use the URL below in plugin manager and let me know if you are seeing the stutters like before.

https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.21.zip

[cbarth3]

Sounds good. I have a print finishing up late tonight so I will try it out in the morning.

[cbarth3]

Tried to make some videos to review. With just the plugin enabled it still has short little pauses. https://photos.app.goo.gl/xe5Mjh1rkTBynfX89

[jneilliii]

Thanks, and I understand printing the same file with tplink disabled that doesn't happen. Can you please share your gcode file, your tplink settings (including the plug settings), and key me know what firmware/ version your printer is running? Any reason you've not upgraded octoprint to the latest?

[cbarth3]

Originally that was a fresh install of octopi. I have since updated to the latest. The videos I uploaded it was on the latest of octoprint. I don't have the plugin enabled right now I will have to get that over a little later. Printer is Ender 3 pro v1.1.6.2 Ender3_Level_Calibration_circles.txt

[cbarth3]

Here are the settings screens for the plugin

[debren27]

Hi @jneilliii, thanks for pointing me toward this. I'm definitely seeing the stuttering issue and have narrowed it down to the TPLink plugin.

Prints are smooth from SD card of course, but also with Octoprint in Safe Mode, Octoprint in normal mode with no plugins enabled, and even with some plugins enabled. Granted, it's a little slower under each of these successive combinations, but it's only when I enable OctoPrint-TPLinkSmartplug that I see true stuttering -- actual stopping -- and therefore blobs and zits from ooze. The difference between no plugins enabled, and only TPLink enabled, is stark.

Besides the stuttering, I also noticed that the octoprint process used 100% of the cpu when only TPLink was enabled, versus 50% when no plugins were enabled.

I'm happy to provide whatever info I can, and to do testing. I don't have videos yet but I can make some if needed.

My setup: OctoPrint 1.4.0 running on OctoPi 0.15.1 on an RPi 3B+ OctoPrint-TPLinkSmartplug 0.9.23 CR-10S S5 TH3D EZBoard on latest TH3D Unified (Marlin) firmware Kasa Smart Plug 6-Outlet Power Strip HS300 http://amzn.com/B07G95FFN3 Kasa Smart Plug by TP-Link HS100 http://amzn.com/B0178IC734

The Kasa/TP-Link stripe does appear to have energy monitoring, although I don't use it. The other plug does not have energy monitoring. I have 3 of the plugs on the stripe, plus the other plug for the bed heater, configured in the plugin. (I plug the strip into a UPS but can't run the 1300W Keenovo through the UPS.)

My testing all weekend was on a 5-minute print (grid of 200mm-around 2-layer rings with segment sizes of .1mm to 2mm in increments of .1), which I now know exacerbates the issue based on this thread.

Let me know next steps.

Here's my plugin config:

Here's my gcode: curve stutter test v2 4x5 ringCircumf120 fs.1-2 60mmps horizExpNo.gcode.txt

Here's the debug log: plugin_tplinksmartplug_debug.log.gz

[jneilliii]

@debren27 to confirm my suspicions could you remove all plugs except the HS100 and try the quick print again? I think you are right, the small print would exaggerate the issue. Also, what type of device are you running OctoPrint on and do you happen to know what your firmware's buffer size is configured at?

[debren27]

@debren27 to confirm my suspicions could you remove all plugs except the HS100 and try the quick print again? I think you are right, the small print would exaggerate the issue. Also, what type of device are you running OctoPrint on and do you happen to know what your firmware's buffer size is configured at?

I should be able to do this tonight. I have an 8-hour print running right now but after that.

Just to confirm, I'll remove the 3 HS300 plugs (or maybe just all of them) from the plugin configuration, but leave the plugin enabled. (I'll try to find a config file to back up so I don't have to totally reconfig afterward. :) ) Leaving the printer actually plugged into the smart outlets shouldn't matter and should be fine.

I have a Pi 3B+.

Buffer sizes are set by TH3D but I believe they increase them beyond the default. BUFSIZE is 32 if that's the right var. https://github.com/houseofbugs/TH3D-Unified-U2.R1/blob/master/Marlin/Configuration_adv.h#L388

[jneilliii]

Yeah, possibly having it with no plugs, and then one with just the HS100 setup. Specifically interested in the CPU utilization. I know @cp2004 mentioned a potential PR he was going to put together to fix a potential issue with the plugin's threading and that might be at play here as well.

[debren27]

OK I did 4 different tests, all with Octoprint in normal (not Safe) mode and no other 3rd party plugins enabled: no plugins enabled, not even TPLink: printing: https://www.youtube.com/watch?v=3ALnoGthY0E top: https://www.youtube.com/watch?v=cYO3y05t1YM only TPLink plugin with all 4 plugs: printing: https://www.youtube.com/watch?v=qucx70YYv8A top: https://www.youtube.com/watch?v=xHmCgTXXvK8 only TPLink plugin with only HS100 plug: printing: https://youtu.be/TFYE5wmMWPg top: https://youtu.be/0BDoNL1q_TQ only TPLink plugin with no plugs: printing: https://youtu.be/-SDd1uzki_s top: https://youtu.be/34F0xDmCQA0

All the videos of top (CPU usage) were taken right after the video of printing.

As we pretty much expected, the biggest difference was the HS300 plugs. Again CPU usage went from 50% to 100% with the plugs configured.

Here's the debug log, obviously only for 3 of the 4 prints (in the order above): plugin_tplinksmartplug_debug.log.2020-07-27.gz

And here's a picture of the resultant blobbing in the test with all 4 plugs still configured:

[jneilliii]

Thank you so much @debren27 I will be looking into this issue more in depth this weekend. I assume I'm not effected by this issue like you guys because I'm running a Pi4 4GB model. I can see it would definitely be exaggerated on circular round things if you're not using arc support in your firmware and the Arc Welder plugin. It's all the little moves that exaggerate the issue.

[jneilliii]

@debren27 all your videos are marked private, so I can't open them...

[debren27]

@debren27 all your videos are marked private, so I can't open them...

Should be fixed.

I just ordered an RPi 4B 8GB to help with this (before I knew about this issue) so I may get a chance to test that. Somehow I didn’t know about Arc Welder so I’ll have to look into that.

[jneilliii]

If your firmware supports arcs it's a miracle plugin.

[jneilliii]

Videos are very helpful in seeing the problem. It does appear that the HS100 plug alone is unaffected by the issue, which does lead me to believe it's the print progress polling for energy data. I will be checking into this over the weekend to see if I can find a solution.

[cp2004]

@debren27 Just a warning, the 8GB pi 4 is not currently officially supported but you can head to the forum to find a beta 64bit image, or I believe (Might be wrong) the 0.18 nightly builds support it.

Been following your issue, will look a little closer into it at some point.

[debren27]

@debren27 Just a warning, the 8GB pi 4 is not currently officially supported but you can head to the forum to find a beta 64bit image, or I believe (Might be wrong) the 0.18 nightly builds support it.

Oh thanks. Last I knew the 4 wasn't supported at all but then I read that it is now. Didn't realize the 8GB was different. But now that I know my stuttering is an issue with the plugin and my plugs, I have less reason to upgrade, so maybe I'll wait.

Been following your issue, will look a little closer into it at some point.

I really appreciate it!

[jneilliii]

So far at this point we don't think there is anything that would be computationally intense from the plugin side. From what I'm seeing though, the performance without the energy monitoring running seems to be "normal". Can't say for sure, because it looks like the top video for just the one plug isn't linked.

In any case I think as a test I'll remove the print progress processing from the equation by making energy monitoring an option in settings to avoid the double hit to the plug status check, etc. I'm fairly confident in taking that out it will help your situation, at the loss of not having the power graphing during a print.

I wonder if it might have something to do with the logging to sqlite3 during the process maybe @cp2004?

[cp2004]

That did seem to be the only thing I spotted that could cause a 'stutter' in the print, might have to investigate. @debren27 Do you know what class/speed your SD card is? Could be a source of slowness, potentially? The 3B+ should have no problem with this, I used to use a 3B - until it died :( (separate story) but just so I'm clear you have 3 plugs that have energy monitoring (Even if you don't use it, I think the plugin will log it) and one without?

@jneilliii Good call on the sqlite db - I have used sqlalchemy with sqlite db, but left it connected the whole way through. The query itself shouldn't be hard, however it may be the connect/query/disconnect (x3) cycle that slows things down. Best tactic I think here is to isolate it and test how long that takes.

[debren27]

because it looks like the top video for just the one plug isn't linked.

Argh, sorry, fixed above in my original comment.

[debren27]

Also, I really like the idea of making energy monitoring optional, at least as a stopgap. For me the auto-shutoff is the killer app of this plugin. I love energy monitoring but I can live without it, to have both print quality and auto-shutoff.

Do you know what class/speed your SD card is? Could be a source of slowness, potentially?

I believe it's this Samsung EVO Select 128GB Class 10 U3: http://amzn.com/B06XWZWYVP

The 3B+ should have no problem with this, I used to use a 3B - until it died :( (separate story) but just so I'm clear you have 3 plugs that have energy monitoring (Even if you don't use it, I think the plugin will log it) and one without?

That's correct about the plugs. And I agree about the 3B+, I thought all indications were that it was more than qualified for Octoprint.

it may be the connect/query/disconnect (x3) cycle that slows things down.

Based on my 30 years of sysadmin experience, that sounds very plausible! :)

[jneilliii]

it may be the connect/query/disconnect (x3) cycle that slows things down.

Based on my 30 years of sysadmin experience, that sounds very plausible! :)

I kind of agree now that I look at the code again. I should just create the connection on startup and leave it open until shutdown maybe? That could definitely take some cycles out of the equation. Just not sure how to handle a potential disconnected state for some reason.

[jneilliii]

@debren27 I've added a new branch to begin working on this issue and have the optional print progress polling in the settings now. If you want to give it a try just install using the URL below in Plugin Manager > Get More > ...from URL and click Install button. Not need to uninstall the previous version or anything. The new option is disabled by default so after installing it should be ready to give it a test.

https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip

[jneilliii]

@cp2004 what's a good method of timing during debugging in Pycharm?

[debren27]

Something's strange because I install from the 0.9.24 URL and it looks like it's doing it but then it says it installed 0.9.23. Not sure if that matters. I'll include the output below.

Is "Enable polling during print progress" the new option? I do see that.

I restored arrSmartplugs from a backup of config.yaml. I hope that's ok.

Here's the test print: printing: https://youtu.be/q-ISD7W_Bls top: https://youtu.be/bRWw4bBceuQ

Looks great! This makes me very happy, because I can have the on/off/autooff functionality AND the print quality. All I have to sacrifice is energy monitoring which I don't use so it's like giving up broccoli for Lent. And hopefully you'll fix that for those that want it (and maybe I'll want it later).

I checked the debug log and there's absolutely nothing from when it was printing. The print started at 22:49 and the log jumps from 22:39 to 22:55, which is when I canceled the print.

Here's my installer window output:

Installing plugin from https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip...
/home/pi/oprint/bin/python2 -m pip --disable-pip-version-check install https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip --no-cache-dir
Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple
Collecting https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip
Downloading https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip (2.4 MB)
Requirement already satisfied (use --upgrade to upgrade): OctoPrint-TPLinkSmartplug==0.9.23 from https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip in /home/pi/oprint/lib/python2.7/site-packages
Requirement already satisfied: OctoPrint in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint-TPLinkSmartplug==0.9.23) (1.4.0)
Requirement already satisfied: flask<0.13,>=0.12 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.12.5)
Requirement already satisfied: Jinja2<2.9,>=2.8.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2.8.1)
Requirement already satisfied: tornado==4.5.3 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (4.5.3)
Requirement already satisfied: regex!=2018.11.6 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2018.2.21)
Requirement already satisfied: Flask-Login<0.5,>=0.4.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.4.1)
Requirement already satisfied: Flask-Babel<0.13,>=0.12 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.12.2)
Requirement already satisfied: Flask-Assets<0.13,>=0.12 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.12)
Requirement already satisfied: blinker<1.5,>=1.4 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.4)
Requirement already satisfied: werkzeug<0.17,>=0.16 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.16.1)
Requirement already satisfied: cachelib<0.2,>=0.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.1)
Requirement already satisfied: PyYAML<6,>=5.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (5.1)
Requirement already satisfied: markdown<3.2,>=3.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.1.1)
Requirement already satisfied: pyserial<3.5,>=3.4 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.4)
Requirement already satisfied: netaddr<0.8,>=0.7.19 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.7.19)
Requirement already satisfied: watchdog<0.10,>=0.9.0 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.9.0)
Requirement already satisfied: sarge==0.1.5post0 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.1.5.post0)
Requirement already satisfied: netifaces<0.11,>=0.10.9 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.10.9)
Requirement already satisfied: pylru<1.3,>=1.2 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.2.0)
Requirement already satisfied: rsa<5,>=4.0 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (4.0)
Requirement already satisfied: pkginfo<1.6,>=1.5.0.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.5.0.1)
Requirement already satisfied: requests<3,>=2.22.0 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2.23.0)
Requirement already satisfied: semantic_version<2.9,>=2.8 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2.8.4)
Requirement already satisfied: psutil<5.7,>=5.6.5 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (5.6.7)
Requirement already satisfied: Click<8,>=7 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (7.0)
Requirement already satisfied: awesome-slugify<1.7,>=1.6.5 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.6.5)
Requirement already satisfied: feedparser<5.3,>=5.2.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (5.2.1)
Requirement already satisfied: future<0.19,>=0.18.2 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.18.2)
Requirement already satisfied: websocket-client<0.57,>=0.56 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.56.0)
Requirement already satisfied: wrapt<1.12,>=1.11.2 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.11.2)
Requirement already satisfied: emoji<0.6,>=0.5.4 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.5.4)
Requirement already satisfied: frozendict<1.3,>=1.2 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.2)
Requirement already satisfied: sentry-sdk==0.13.2 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.13.2)
Requirement already satisfied: filetype<2,>=1.0.5 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.0.5)
Requirement already satisfied: futures<3.4,>=3.3 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.3.0)
Requirement already satisfied: monotonic<1.6,>=1.5 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.5)
Requirement already satisfied: scandir<1.11,>=1.10 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.10.0)
Requirement already satisfied: chainmap<1.1,>=1.0.3 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.0.3)
Requirement already satisfied: typing<4,>=3.7.4.1 in /home/pi/oprint/lib/python2.7/site-packages (from OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.7.4.1)
Requirement already satisfied: itsdangerous>=0.21 in /home/pi/oprint/lib/python2.7/site-packages (from flask<0.13,>=0.12->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.24)
Requirement already satisfied: MarkupSafe in /home/pi/oprint/lib/python2.7/site-packages (from Jinja2<2.9,>=2.8.1->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.0)
Requirement already satisfied: singledispatch in /home/pi/oprint/lib/python2.7/site-packages (from tornado==4.5.3->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.4.0.3)
Requirement already satisfied: certifi in /home/pi/oprint/lib/python2.7/site-packages (from tornado==4.5.3->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2018.4.16)
Requirement already satisfied: backports-abc>=0.4 in /home/pi/oprint/lib/python2.7/site-packages (from tornado==4.5.3->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.5)
Requirement already satisfied: Babel>=2.3 in /home/pi/oprint/lib/python2.7/site-packages (from Flask-Babel<0.13,>=0.12->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2.5.3)
Requirement already satisfied: webassets>=0.11.1 in /home/pi/oprint/lib/python2.7/site-packages (from Flask-Assets<0.13,>=0.12->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.12.1)
Requirement already satisfied: setuptools>=36 in /home/pi/oprint/lib/python2.7/site-packages (from markdown<3.2,>=3.1->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (39.1.0)
Requirement already satisfied: argh>=0.24.1 in /home/pi/oprint/lib/python2.7/site-packages (from watchdog<0.10,>=0.9.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.26.2)
Requirement already satisfied: pathtools>=0.1.1 in /home/pi/oprint/lib/python2.7/site-packages (from watchdog<0.10,>=0.9.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.1.2)
Requirement already satisfied: pyasn1>=0.1.3 in /home/pi/oprint/lib/python2.7/site-packages (from rsa<5,>=4.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.4.2)
Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /home/pi/oprint/lib/python2.7/site-packages (from requests<3,>=2.22.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.22)
Requirement already satisfied: chardet<4,>=3.0.2 in /home/pi/oprint/lib/python2.7/site-packages (from requests<3,>=2.22.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (3.0.4)
Requirement already satisfied: idna<3,>=2.5 in /home/pi/oprint/lib/python2.7/site-packages (from requests<3,>=2.22.0->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2.6)
Requirement already satisfied: Unidecode<0.05,>=0.04.14 in /home/pi/oprint/lib/python2.7/site-packages (from awesome-slugify<1.7,>=1.6.5->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (0.4.21)
Requirement already satisfied: six in /home/pi/oprint/lib/python2.7/site-packages (from websocket-client<0.57,>=0.56->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (1.11.0)
Requirement already satisfied: pytz>=0a in /home/pi/oprint/lib/python2.7/site-packages (from Babel>=2.3->Flask-Babel<0.13,>=0.12->OctoPrint->OctoPrint-TPLinkSmartplug==0.9.23) (2018.4)
Building wheels for collected packages: OctoPrint-TPLinkSmartplug
Building wheel for OctoPrint-TPLinkSmartplug (setup.py): started
Building wheel for OctoPrint-TPLinkSmartplug (setup.py): finished with status 'done'
Created wheel for OctoPrint-TPLinkSmartplug: filename=OctoPrint_TPLinkSmartplug-0.9.23-py2-none-any.whl size=2221485 sha256=6de3adcce5640116342c031e6d156d3e854709b8555ec14ec2f41520ab3c87db
Stored in directory: /tmp/pip-ephem-wheel-cache-boFBSz/wheels/09/32/b0/ca7140508d9f961fb63bfe26520cf47b851646ba7fa4e498cf
Successfully built OctoPrint-TPLinkSmartplug
Looks like the plugin was already installed. Forcing a reinstall.
/home/pi/oprint/bin/python2 -m pip --disable-pip-version-check install https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip --no-cache-dir --ignore-installed --force-reinstall --no-deps
Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple
Collecting https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip
Downloading https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip (2.4 MB)
Building wheels for collected packages: OctoPrint-TPLinkSmartplug
Building wheel for OctoPrint-TPLinkSmartplug (setup.py): started
Building wheel for OctoPrint-TPLinkSmartplug (setup.py): finished with status 'done'
Created wheel for OctoPrint-TPLinkSmartplug: filename=OctoPrint_TPLinkSmartplug-0.9.23-py2-none-any.whl size=2221485 sha256=f1e6d5eaf8e03d5d080ee650c0ad64bd2f68feacf23b1e0f8f1cbe3fef883ec2
Stored in directory: /tmp/pip-ephem-wheel-cache-RhZexO/wheels/09/32/b0/ca7140508d9f961fb63bfe26520cf47b851646ba7fa4e498cf
Successfully built OctoPrint-TPLinkSmartplug
Installing collected packages: OctoPrint-TPLinkSmartplug
Successfully installed OctoPrint-TPLinkSmartplug-0.9.23
Done!

[cp2004]

@cp2004 what's a good method of timing during debugging in Pycharm?

@jneilliii Usually I just add some points where it saves the time, then calculate. Obviously if you are going down to the wire, printing times takes slightly longer (But incredibly incredibly small). This is what I did last night https://gist.github.com/cp2004/8ff1d64254b05e67b929728d5f7a66ba

One more thing to bear in mind, PyCharm's debugger will run a lot slower than in real life, but would still be an ok representation.

On my Windows machine, it took 0.06 secs but I have a fast SSD etc, haven't had a chance to put it on a Pi and record the output. Based on it taking 0.06, it could take in the region of 0.2secs or more on the Pi, which could be enough to stutter OctoPrint.

[debren27]

I kinda doubt it's related, but now I'm noticing that the plugin often turns off my HS100 for no apparent reason. I may have seen this once or twice before I upgraded tho. If I can narrow down why and it's not a boneheaded setting of mine, I'll post here or open a new issue.

[jneilliii]

I have seen cases of back-powered printers from USB connection to pi cause this if you have disconnect monitoring enabled.

[jneilliii]

@debren27 just merged in changes from @cp2004 on the timers and was hoping you'd be willing to test to see if with the print progress polling enabled if this new update still does the same thing or not? You can install using the URL below in Plugin Manager > Get More > ...fromURL

https://github.com/jneilliii/OctoPrint-TPLinkSmartplug/archive/0.9.24.zip

[debren27]

Sorry, I wish I had better news, but I still get a lot of stuttering, and 100+% CPU. I'll post the videos after work.

I did upgrade Octoprint to 1.4.2 in the meantime, but I wouldn't imagine that would make much difference. I first did a confirmation test by enabling polling in the old version of your plugin (and left all my other plugins still enabled) and got really bad stuttering. Probably other plugins made it a little worse.

Then I upgraded by copy and pasting your URL (although the installer output still said 0.9.23), disabled all other plugins, and rebooted. I'm printing the exact same gcode file as before.

[jneilliii]

That's unfortunate news. I was hoping those changes would have cleared up some of that load. No need for videos, I'll have to dig in more on this. In the meantime I'm going to go ahead and release this new version that resolves another issue.

[debren27]

Here's the plugin manager log for the install, and the debug log for the print. plugin_tplinksmartplug_debug.log plugin_pluginmanager_console.log

[cp2004]

@debren27 I'm going to take a further look into this, but just want a couple of things:-

• Is it constant 100% CPU usage that you are seeing, and is it with or without polling enabled?
• Does it only happen with more than one plug, or is it solved with only a single plug set up.

Thanks for your consistent help with this!

[debren27]

• Is it constant 100% CPU usage that you are seeing, and is it with or without polling enabled?

With polling enabled, watching top for half a minute, it bounces between 90% and 110%. So it's not quite pegged at a constant 100%, but it averages around that. But I believe it stays that high for the entire print, especially since it's a small (10-min) print consisting entirely of small circles with lots of segments.

Without polling enabled, while I didn't test the new version, the previous version bounced between 45% and 60%. That's this video: https://youtu.be/bRWw4bBceuQ. I can test the new version with polling disabled if you'd like.

• Does it only happen with more than one plug, or is it solved with only a single plug set up.

Do you mean only a single plug with energy monitoring, or only the single HS100 (which doesn't have energy monitoring) that I left enabled for some of the previous tests? I've only tested the new plugin with all 4 plugs (3 with EM) enabled. I can pare that down to one plug with EM if you'd like.

Thanks for your consistent help with this!

My pleasure. I appreciate the fast responses from you both. And it's kinda cool to be involved!

Let me know what other tests you'd like me to run.

[cp2004]

I might have to hook it up to a debugger and run some hard prints. I've found a bearing cover that actually broke the virtual printer because it had too many segments, I reckon I can re-create something like what you're seeing.

Do you mean only a single plug with energy monitoring, or only the single HS100 (which doesn't have energy monitoring) that I left enabled for some of the previous tests? I've only tested the new plugin with all 4 plugs (3 with EM) enabled. I can pare that down to one plug with EM if you'd like.

I've only got one EM plug here, so that's what I was wondering, as I don't have this issue - but then I also have a Pi 4, where they increased the speed of the SD card interface over the 3. Might see if I can get something to run on a Pi Zero, to really restrict the resources...

Nothing more for now, I think I've got a good picture of the issue, and (hopefully) an idea to investigate

[jneilliii]

You both are great. I really appreciate the help @cp2004 and great support interaction with @debren27. It's always good to get someone else to look at this with fresh eyes. I too am running a pi4 4gb model with no issues, but do have multiple plugs with energy monitoring and have never seen this.

[debren27]

I bought a Pi 4B (actually two; first the 8GB then you gave me the heads up about compatibility, so I got a 4GB) and I finally imaged the SD card over the weekend but "luckily" for you both I haven't deployed it yet so the 3B+ is still driving my printer. I have plenty of other projects and turning off polling is a great workaround for me so I won't rush to deploy the upgrade.