-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Tuesday, 2019-12-31 at 19:00 -0500, Felix Miata wrote:
Carlos E. R. composed on 2020-01-01 00:06 (UTC+0100):
Almost two gigabytes.
How many mozilla app instances is that?
One.
Makes 2G usage seem exorbitant.
I'm used to it. This moment, just out of hibernation half an hour ago - top output sorted by RES memory. Using Alpine to post, so no line wrapping issues. top - 13:40:59 up 23:47, 2 users, load average: 0,19, 0,17, 0,26 Tasks: 487 total, 1 running, 485 sleeping, 0 stopped, 1 zombie %Cpu(s): 6,0 us, 1,0 sy, 0,0 ni, 91,6 id, 1,4 wa, 0,0 hi, 0,0 si, 0,0 st KiB Mem : 8161256 total, 3064304 free, 2874280 used, 2222672 buff/cache KiB Swap: 25165820 total, 21942148 free, 3223672 used. 4752628 avail Mem PID USER PR NI VIRT RES SHR SWAP S %CPU %MEM TIME+ COMMAND 6264 cer 20 0 3955540 666708 109840 33980 S 0,595 8,169 10:13.91 thunderbird-bin 8857 cer 20 0 3807704 465564 168852 23732 S 0,893 5,705 5:02.12 firefox 9064 cer 20 0 2971416 286368 68080 0 S 0,000 3,509 0:54.66 Web Content 9040 cer 20 0 3130720 205592 58344 0 S 0,298 2,519 1:53.03 Web Content 9015 cer 20 0 2795016 193968 61768 0 S 0,000 2,377 0:11.39 Web Content 4034 root 20 0 540556 142908 117732 41600 S 13,10 1,751 1:55.62 X 9182 cer 20 0 26,677g 139920 27596 0 S 0,298 1,714 0:17.70 WebExtensions 8997 cer 20 0 2826632 137252 52628 0 S 0,000 1,682 4:26.05 Web Content 9112 cer 20 0 2804980 134136 28080 0 S 0,000 1,644 0:22.47 Web Content 21830 cer 20 0 2710312 122972 68568 0 S 0,893 1,507 0:51.11 Web Content 9088 cer 20 0 2720932 117396 37332 0 S 0,000 1,438 0:07.34 Web Content 8959 cer 20 0 2736492 105016 38048 16796 S 0,000 1,287 0:10.73 Web Content 8656 cer 20 0 4063364 85376 12124 76852 S 0,298 1,046 0:14.51 java Just look at WebExtensions, 26 GB of virtual memory! I'm told VIRT does not matter. 465564+286368+205592+193968+139920+137252+134136+122972+117396+105016=1908184 bytes. It is what it is... Firefox is a memory hog.
How many tabs? How many windows?
A few, most of them not loaded.
As in less than 10? Seems more exorbitant.
I'll count them. 22+24+9+7+5+4+9=80 tabs in 7 windows. Not even a hundred. And most of the windows have only one tab actually loaded - they don't load till one clicks on the tab to view it. Or at least, they are not rendered.
How old is the history?
Not much. This profile I created a month or two ago. Why would history matter?
It probably should matter de minimus in a young profile, but history does get a lot of edits.
History is just a text file. Can't be a megabyte.
I don't think 2GB is too much if the tab count and history are huge.
Well, it is a high percent of the machine. Actually, I created a new FF profile to start again with minimal tab usage. Maybe less than a hundred. No way to count them, except manually.
https://addons.mozilla.org/en-US/firefox/addon/tab-counter-webext/ indicates otherwise.
Ah, so there is an addon to count them. [...] Says there are 7 windows and 80 tabs.
What may make a difference is whether builds are static or not. All running in my previous post are Mozilla's static builds. I have 12 Firefox profiles and as many installed versions. My ability to utilize openSUSE's Mozilla rpms is severely limited. I couldn't be closing one version in order to open another even if I could have multiple rpms installed at once.
I use the official openSUSE rpm, as always.
That may contribute to or be the problem. Mozilla RAM usage rarely takes any of my time, while I see others regardless of distro complain about it more than a little. The difference between is I rarely run any but static builds. What build types are Firefox, TB and Rust developers using? Maybe it's mainly mere mortal users who "test" .rpms and .debs, and/or devs don't keep any running long enough to gauge what happens in release world usage.
Static build uses more memory than shared.
http://download.opensuse.org/repositories/mozilla:/alpha/ & http://download.opensuse.org/repositories/mozilla:/beta/ are void of 15.1 directories.
You can test a Mozilla build of same version using the exact same profile without corrupting it in any way. I've gone both ways lots of times over the years, though not in a while. I wouldn't expect the safety to have been affected. Even if it was, backups can be restored. ;-)
- -- Cheers, Carlos E. R. (from openSUSE 15.1 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCXgyXXBwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVNdcAn3U0tYvLSrrLXhUaYx5B kG25KJvRAJ4v9FeuSm+4MlyQX1uHvI405i34+g== =tyFi -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org