Significant performance degradation after upgrade to v28

Users and developers helping users with technical Pale Moon issues (Windows and other non-Linux O.S.). Please direct questions about the Linux version to the appropriate Linux board.

Moderators: trava90, satrow

Forum rules
This board is for technical/usage questions and troubleshooting for the Pale Moon browser only. The main focus here is on Pale Moon on Windows. Please direct your questions for Linux, Android and Mac to the dedicated boards.
Technical issues and questions not related to the Pale Moon browser should be posted in "technical chat"
Please keep off-topic and general discussion out of this board, thank you!
User avatar
GogrillaMincefriend
Newbie
Newbie
Posts: 6
Joined: 2019-01-10, 09:22

Re: Significant performance degradation after upgrade to v28

Unread post by GogrillaMincefriend » 2019-01-11, 00:33

Some more digging on my side regarding the memory issue; like I said, it felt a lot like the process was swapping a lot, and a quick look at resource monitor showed regular writes out to the page file, and perfmon showed that the page file was 100% full - even though I only use a 1GB page file, this never happens (as I have 32GB of RAM and a regular commit of 12-16GB so page file practically never gets used).

Firing up procmon to monitor C:\pagefile.sys (first "Filter > Enable advanced output", then "Filter > Filter... > Path is C:\pagefile.sys" to show reads and writes from pagefile) there's a write to pagefile from System (showing as IRP_MJ_WRITE) almost exactly every second, and the only process to ever read from the pagefile in the last 15mins has been palemoon - as depicted in Yet Another Attachment.

I don't know if there are any tools available to verify whether or not its PM that is actually responsible for this activity since TTBOMK windows doesn't provide any easy access to this sort of thing. It's dead easy on linux - assuming only a single instance of palemoon running (it'll fail if there are multiple PIDs) you can just run the following command to report how much has been swapped:

Code: Select all

root@krikkit:~# grep VmSwap /proc/`pidof palemoon`/status
VmSwap:      848 kB
(I say this only to demonstrate the likelihood of why I'm not seeing this issue on my linux systems despite identical extensions and browsing profile)

I've disabled my page file in the anticipation that if PM is swapping out, killing the page file will prevent this behaviour; but this'll require a reboot, and I'm likely weeks away from another reboot (barring crashes or hardware gremlins) so if anyone else is able to entertain firing up procmon to see if they've got anything resembling the same issue (and maybe also think about disabling their page file if they've got enough RAM to play with and don't mind a reboot and potential problems due to lack of page file) then I'd love to hear back from someone.

Thanks for the kind words CodeLurker - I'm definitely no guru (and I don't want to cause a ruckus for what might be an issue unique to my system and throwing red herring's everywhere) but I hope some other people might find the info useful. As you say, sporadic bugs are the worst (I had one at work that took 3yrs to get to the bottom of because it only occurred every 3 months or so and I'm happy to report that both the vendor and Microsoft were both entirely useless in assisting with debugging it).
Attachments
palemoon_procmon_pagefile_activity.png
page file activity - writes every 1s, and reads only from palemoon.exe

Post Reply