user20C Posted March 27, 2020 Share Posted March 27, 2020 Hey everyone, I was hoping to get some help with this memory issue I am having. About every 24 hours I have to reboot my server because the memory keeps maxing out. when I run htop in the terminal and sort by mem % i get the attached picture. It has been going on for a few weeks. I haven't changed anything hardware wise with the server and I don't recall any unraid updates triggering it. It happened with 6.8.2 and 6.8.3. Attached are a screenshot of the terminal window with htop running and my diagnostics. Any help would be greatly appreciated. tower-diagnostics-20200327-1622.zip Quote Link to comment
user20C Posted March 29, 2020 Author Share Posted March 29, 2020 Still doing the daily restarts. Last night I let the server keep running until shfs filling crashed. At that point ram % went down to 3% but all my shares disappeared. A restart brought everything back. Can anyone tell me what shfs is or does? Is there any way to tell if a certain plugin or docker is using it? I have disabled cache directories based on some forum searches but that doesn't help. A forum search didn't really bring up anything, most other problems with shfs have dealt with it utilizing to much CPU which is not happening for me. Any help would be greatly appreciated! Quote Link to comment
JorgeB Posted March 29, 2020 Share Posted March 29, 2020 Best bet would be to run Unraid without with all dockers disable, if there's no problem like that enable one them by one and keep monitoring until you find the culprit. Quote Link to comment
user20C Posted March 29, 2020 Author Share Posted March 29, 2020 I have disabled docker completely and restarted twice now and shfs is still slowly consuming memory. I guess I need to start with plugins next. Is there switch to shut off all plugins? Poking around a few of them there doesn't seem to be an explicit on/off toggle for them. I guess I will have to uninstall one by one until those processes that are consuming the memory go away. Quote Link to comment
itimpi Posted March 29, 2020 Share Posted March 29, 2020 2 minutes ago, user20C said: I have disabled docker completely and restarted twice now and shfs is still slowly consuming memory. I guess I need to start with plugins next. Is there switch to shut off all plugins? Poking around a few of them there doesn't seem to be an explicit on/off toggle for them. I guess I will have to uninstall one by one until those processes that are consuming the memory go away. Booting in Safe Mode stops all plugins from being started. An alternative approach is to rename the .plg files in the config folder on the flash drive to have a different extension and reboot which will also stop those plugins from being started. Quote Link to comment
user20C Posted March 29, 2020 Author Share Posted March 29, 2020 booted into safe mode with web GUI. With the array not started everything ran as expected. System idling with shfs not showing up. As soon as I started they array, shfs processes started and ram utilization has been slowing going up. I know shfs is a normal process and is expected to run. However i have multiple instances of it and they all slowing eat up the ram. Again this is in safe mode with web GUI enabled. I am observing all this by opening a terminal window and running htop and sorting by mem%. I assume this is an acceptable way to monitor this. So it seems I have an underlying OS problem? Or could a hardware issue cause this. There are alot of ram modules in this server. I purchased it off ebay as is. 192GB ram total. I haven't been in the case recently or changed anything hardware wise. Just thinking out loud here and trying to give as much relevent information as possible. Quote Link to comment
JorgeB Posted March 30, 2020 Share Posted March 30, 2020 Try redoing you flash drive, backup config folder, redo flash, restore only super.dat and the key file, both on the /config folder, see it it still goes up like that. Quote Link to comment
user20C Posted March 30, 2020 Author Share Posted March 30, 2020 (edited) 10 hours ago, johnnie.black said: Try redoing you flash drive, backup config folder, redo flash, restore only super.dat and the key file, both on the /config folder, see it it still goes up like that. I shutdown the server, pulled the USB and put it in my windows PC. I copied the config file to my desktop then formatted the drive. I downloaded the USB creator tool from the unraid website and created a new USB with 6.8.3. I then copied over the super.dat file and the key file from the old config folder to the new config folder. Put the USB back in my server and started it up. After starting the array manually I get the same behavior. Several instances of shfs are running again and slowly consuming memory. Its only been running for 40 mins so total mem% is still low but the trend is there as it was before. So we have rulled out plugins, dockers, and the USB drive. It only happens when the array is started. Another issue I have discovered is if I try to stop the Array the system hangs. I get a message in the bottom left of the browser that says "Array Stopping Retry unmounting user share(s)...) I guess this related to the shfs processes not stopping or hanging. Edited March 30, 2020 by user20C typo Quote Link to comment
JorgeB Posted March 30, 2020 Share Posted March 30, 2020 How come you have Emby running? Quote Link to comment
user20C Posted March 30, 2020 Author Share Posted March 30, 2020 I guess I thought we had already ruled anything docker and plugin related out as the cause. I will restart with docker set to off and see if that changes anything with the newly created USB. Quote Link to comment
user20C Posted March 31, 2020 Author Share Posted March 31, 2020 Ok, I restarted with docker set to off, no plugins installed, and no VMs. After starting the array i saw 1.8G of memory being used. 30 minutes in it was at 4.59G. 40 mins it was at 6.06G. And at 1 hour it was at 8.36G of memory used. So a steady increase. I beleive this would go on for about 24 hours until it maxes out the total 189G available. I attached a screenshot of htop again sorted by mem% and also added another diagnostics file. Any ideas or help is greatly appreciated and thank you for the help so far. tower-diagnostics-20200330-2034.zip Quote Link to comment
JorgeB Posted March 31, 2020 Share Posted March 31, 2020 Very strange, it's obviously not a general problem since nobody else appears to have this issue but can't see what's causing it for you, please create a bug report for this, mention that this is a completely clean OS install (using existing data disks) and include these latest diags. Quote Link to comment
user20C Posted March 31, 2020 Author Share Posted March 31, 2020 I will, thank you for help provided so far! Quote Link to comment
ecnal.magnus Posted April 16, 2022 Share Posted April 16, 2022 I am having this exact problem, and I started the attached thread in hopes of figuring it out. Did you ever get this corrected? Did it just go away on its own? Is there anything you can tell me that you haven't already posted here? Any help or insight would be greatly appreciated. Quote Link to comment
ecnal.magnus Posted April 16, 2022 Share Posted April 16, 2022 Never mind, I found your other thread. Stopping SMB right now to see if it helps. 1 Quote Link to comment
Healzangels Posted January 8 Share Posted January 8 Hey seem to be having this same issue! Was wondering if you had found a fix. I am reluctant to stop SMB as I use that for mounting network shares to me NAS/additional media content. Thanks! Quote Link to comment
ecnal.magnus Posted January 13 Share Posted January 13 (edited) I had a corrupt share. I couldn't even bring up the share's configuration page. It would just hang like it was trying to load. Luckily I didn't have any data on the share that was important, so I just deleted it all, deleted the share, and everything went back to normal. It might be possible to move data off of a corrupt share, but I didn't try that. I can't even say that is the issue you are having, only that it ended up being my issue. Hope that helps. Edited January 13 by ecnal.magnus Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.