Closed grahamperrin closed 7 years ago
Above:
… the hardware that I normally use for FreeNAS died … a hunch that the client becomes hungry in some Life Preserver replication failure situations. …
After that failure of the replication target, I removed the one and only replication schedule.
The hogging persists, so it's probably safe to ignore my first hunch and instead, revisit the notion from #128:
… message presentation issue might somehow contribute to the memory-hogging aspect of #126, …
Yesterday (2016-11-05) at 01:01:
15:16, 1.5 GB virtual memory used by the client:
Today:
$ last | less
gjp4 pts/1 :0 Sun 6 Nov 09:49 still logged in
gjp4 pts/0 :0 Sat 5 Nov 15:33 still logged in
boot time Sat 5 Nov 15:23
shutdown time Sat 5 Nov 15:21
gjp4 pts/0 :0 Sat 5 Nov 15:21 - 15:21 (00:00)
gjp4 pts/1 :0 Sat 5 Nov 15:02 - 15:17 (00:15)
gjp4 pts/1 :0 Sat 5 Nov 14:58 - 15:01 (00:02)
gjp4 pts/0 :0 Sat 5 Nov 14:44 - 15:17 (00:33)
gjp4 pts/0 :0 Sat 5 Nov 14:06 - 14:43 (00:36)
boot time Sat 5 Nov 14:03
– after leaving the machine overnight (I left the house around 19:00), this morning I arrived home and found the client using 11.7 G virtual memory, 9,278 M resident on a system with 16 GB RAM:
1 [|||||||||||||||||||||||||||||||||||||||||||||||||||| 41.9%] Tasks: 136, 0 thr; 3 running
2 [||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 81.2%] Load average: 1.98 2.07 2.03
3 [||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 48.7%] Uptime: 18:28:59
4 [|||||||||||||||||||||||||||||||||||||||||||||||| 38.2%]
Mem[||||||||||||||||||||||||||||||||||||||||||||||||| 5.03G/15.8G]
Swp[||||||||||||||||||||||||||||||||||||||||||||||| 2.98G/8.00G]
PID USER PRI NI VIRT RES S CPU% MEM% TIME+ Command
1764 gjp4 20 0 11.7G 9278M S 100. 57.5 18h10:16 /usr/local/bin/sysadm-client -session 1055961660e557000147834951100000036840025_1478359250_160353
2706 gjp4 93 0 4084M 3065M R 79.2 19.0 17h01:16 /usr/local/bin/firefox
1420 root -21 -21 646M 173M S 5.0 1.1 15:39.94 X :0 -auth /var/tmp/.PCDMAuth-42a4f6ce2ac57a2858c1 -nolisten tcp
1690 gjp4 21 0 1171M 125M S 2.1 0.8 41:04.88 kdeinit4: kdeinit4: plasma-desktop
2685 gjp4 21 0 549M 36192 S 1.4 0.2 17:19.12 /usr/local/bin/gnome-system-monitor
1775 gjp4 20 0 568M 32940 S 0.0 0.2 0:36.55 /usr/local/bin/pc-nettray
3429 gjp4 20 0 505M 29136 S 0.5 0.2 1:14.59 kdeinit4: kdeinit4: konsole
1686 gjp4 20 0 763M 25872 S 0.3 0.2 6:10.02 kwin -session 104a946447949a000146019633100000613330000_1478359265_220765
1703 gjp4 20 0 282M 24396 S 0.0 0.1 0:06.86 /usr/local/libexec/mysqld --defaults-file=/usr/home/gjp4/.local/share/akonadi/mysql.conf --datadir=/usr/home/gjp4/.local/share/akonadi/db_data/ --socket=/tmp/akonadi-gjp4.UQBVGz/mysql.socket
1740 gjp4 20 0 756M 14804 S 0.0 0.1 0:06.85 kdeinit4: kdeinit4: krunner
17618 gjp4 20 0 350M 13960 S 0.0 0.1 0:14.06 /usr/local/bin/nautilus --new-window
1693 gjp4 20 0 542M 12560 S 0.0 0.1 0:01.02 /usr/local/bin/knotify4
1759 gjp4 20 0 534M 11696 S 0.0 0.1 0:00.48 kget -session 1055961660e557000147262067000000024430033_1478359250_160320
1701 gjp4 20 0 297M 11616 S 0.0 0.1 0:01.28 akonadiserver
1809 gjp4 20 0 480M 11268 S 0.0 0.1 0:00.49 kdeinit4: kdeinit4: klipper
1683 gjp4 20 0 482M 11024 S 0.0 0.1 0:02.00 kdeinit4: kdeinit4: ksmserver
1757 gjp4 20 0 419M 10828 S 0.0 0.1 0:00.48 /usr/local/bin/kgpg -session 1055961660e557000147280896800000017310026_1478359250_160280
17269 gjp4 20 0 322M 10820 S 0.0 0.1 0:01.74 /usr/local/bin/ksnapshot -caption KSnapshot
1674 gjp4 20 0 473M 10652 S 0.0 0.1 0:00.83 kdeinit4: kdeinit4: kglobalaccel
1730 gjp4 20 0 581M 10464 S 0.0 0.1 0:01.30 /usr/local/bin/akonadi_sendlater_agent --identifier akonadi_sendlater_agent
1726 gjp4 20 0 603M 10396 S 0.0 0.1 0:01.37 /usr/local/bin/akonadi_mailfilter_agent --identifier akonadi_mailfilter_agent
1755 gjp4 52 0 712M 10372 S 0.0 0.1 0:00.63 kdeinit4: kdeinit4: kmix -session 104a946447949a000146150907000000027240026_1478354725_972308
1744 gjp4 29 0 268M 9960 S 0.0 0.1 2:41.84 pc-mixer -session 102ca23629333a000143909649400000063880010_1478359250_159067
1679 gjp4 20 0 351M 9916 S 0.0 0.1 0:01.15 /usr/local/bin/kactivitymanagerd
1798 gjp4 20 0 470M 9860 S 0.0 0.1 0:00.68 /usr/local/bin/korgac --icon korgac
1728 gjp4 20 0 422M 9776 S 0.0 0.1 0:01.17 /usr/local/bin/akonadi_newmailnotifier_agent --identifier akonadi_newmailnotifier_agent
1729 gjp4 20 0 527M 9520 S 0.0 0.1 0:01.29 /usr/local/bin/akonadi_notes_agent --identifier akonadi_notes_agent
1719 gjp4 20 0 603M 9492 S 0.0 0.1 0:01.44 /usr/local/bin/akonadi_archivemail_agent --identifier akonadi_archivemail_agent
1794 gjp4 20 0 330M 9260 S 0.0 0.1 0:00.84 /usr/local/lib/kde4/libexec/polkit-kde-authentication-agent-1
1720 gjp4 39 19 353M 9240 S 0.0 0.1 0:01.06 /usr/local/bin/akonadi_baloo_indexer --identifier akonadi_baloo_indexer
1725 gjp4 20 0 379M 8724 S 0.0 0.1 0:01.06 /usr/local/bin/akonadi_maildispatcher_agent --identifier akonadi_maildispatcher_agent
1727 gjp4 20 0 327M 8688 S 0.0 0.1 0:01.06 /usr/local/bin/akonadi_migration_agent --identifier akonadi_migration_agent
1723 gjp4 20 0 267M 8636 S 0.0 0.1 0:00.40 /usr/local/bin/akonadi_agent_launcher akonadi_ical_resource akonadi_ical_resource_0
1689 gjp4 39 19 334M 8620 S 0.0 0.1 0:00.65 /usr/local/bin/baloo_file
1722 gjp4 20 0 334M 8568 S 0.0 0.1 0:01.05 /usr/local/bin/akonadi_followupreminder_agent --identifier akonadi_followupreminder_agent
1442 root 52 0 565M 8568 S 0.0 0.1 0:01.27 /usr/local/bin/PCDM-session -once
1718 gjp4 20 0 273M 8528 S 0.0 0.1 0:00.44 /usr/local/bin/akonadi_agent_launcher akonadi_akonotes_resource akonadi_akonotes_resource_0
1198 root 20 0 141M 8500 S 0.0 0.1 5:09.31 /usr/local/bin/sysadm-binary
1721 gjp4 20 0 339M 8480 S 0.0 0.1 0:01.03 /usr/local/bin/akonadi_birthdays_resource --identifier akonadi_birthdays_resource
1667 gjp4 28 0 687M 8468 S 0.0 0.1 0:01.55 kdeinit4: kdeinit4: kded4
1731 gjp4 20 0 262M 8460 S 0.0 0.1 0:00.40 /usr/local/bin/akonadi_agent_launcher akonadi_vcard_resource akonadi_vcard_resource_0
1663 gjp4 20 0 375M 8440 S 0.0 0.1 0:00.24 kdeinit4: kdeinit4 Running...
1724 gjp4 20 0 271M 8384 S 0.0 0.1 0:00.41 /usr/local/bin/akonadi_agent_launcher akonadi_maildir_resource akonadi_maildir_resource_0
1754 gjp4 29 0 384M 8332 S 0.0 0.1 0:00.03 kdeinit4: kdeinit4: kmix -session 104a946447949a000146150907000000027240026_1478354725_972308
1696 gjp4 20 0 304M 8296 S 0.0 0.1 0:00.56 /usr/local/bin/kuiserver
1681 gjp4 20 0 323M 8200 S 0.0 0.0 0:00.27 /usr/local/lib/kde4/libexec/kdeconnectd
1806 gjp4 36 0 384M 8120 S 0.0 0.0 0:00.02 kdeinit4: kdeinit4: kmix
0 root -16 0 0 8112 S 0.0 0.0 6:33.76 kernel
1734 gjp4 31 0 448M 7780 S 0.0 0.0 0:00.08 kdeinit4: kdeinit4: kmixctrl --restore
6265 gjp4 20 0 25620 6752 R 0.0 0.0 0:00.02 htop
1665 gjp4 20 0 378M 6584 S 0.0 0.0 0:00.18 kdeinit4: kdeinit4: klauncher --fd=8
1799 gjp4 39 19 171M 6376 S 0.0 0.0 0:03.38 /usr/local/libexec/tracker-miner-fs
17658 gjp4 20 0 199M 5256 S 0.0 0.0 0:00.08 /usr/local/bin/mate-screensaver --no-daemon
1748 gjp4 20 0 268M 4212 S 0.0 0.0 0:29.05 /usr/local/bin/pulseaudio --start --log-target=syslog
2705 gjp4 20 0 40476 3892 S 0.0 0.0 7:14.95 libgtop-server
2722 gjp4 20 0 38008 3612 S 0.0 0.0 0:00.22 /usr/local/libexec/dconf-service
1699 gjp4 20 0 102M 3612 S 0.0 0.0 0:01.41 /usr/local/bin/akonadi_control
1800 gjp4 39 19 152M 3548 S 0.0 0.0 0:00.69 /usr/local/libexec/tracker-extract
5864 gjp4 20 0 13956 3408 S 0.0 0.0 0:00.02 /usr/local/bin/bash
2698 gjp4 20 0 64796 3368 S 0.0 0.0 0:00.43 /usr/local/libexec/at-spi2-registryd --use-gnome-session
1803 gjp4 39 19 127M 3364 S 0.0 0.0 0:00.49 /usr/local/libexec/tracker-miner-apps
$ date ; uptime ; freebsd-version ; uname -a
Sun 6 Nov 2016 09:51:55 GMT
9:51a.m. up 18:29, 2 users, load averages: 1.98, 2.07, 2.02
12.0-CURRENT
FreeBSD momh167-gjp4-hpelitebook850g2-trueos.university.brighton.ac.uk 12.0-CURRENT FreeBSD 12.0-CURRENT #14 1717ae3(drm-next-4.7): Fri Oct 28 13:49:49 UTC 2016 root@gauntlet:/usr/obj/usr/src/sys/GENERIC amd64
$
More than half an hour after a SIGTERM kill of the client, the OS still feels clunky (Firefox slow to respond to keystrokes, and so on) –
– and I'll probably not get rid of the clunkiness without a restart of the OS, so this bug is quite disruptive.
(This issue is probably ready for reassignment to a developer; read alongside https://github.com/trueos/trueos-core/issues/128#issuecomment-258767269 where the problem may be reproducible with multiple desktop environments.)
I started the OS and used:
Half-expecting leaps in usage of memory at hourly intervals, shortly after 21:00 I observed a leap.
Around twenty minutes later, no remarkable increase (below) but I half-expect leaps to continue. I'll review usage after 22:00 and/or maybe tomorrow morning.
$ date ; freebsd-version ; uname -a ; htop
Sat 12 Nov 2016 21:22:43 GMT
12.0-CURRENT
FreeBSD momh167-gjp4-hpelitebook850g2-trueos.university.brighton.ac.uk 12.0-CURRENT FreeBSD 12.0-CURRENT #14 1717ae3(drm-next-4.7): Fri Oct 28 13:49:49 UTC 2016 root@gauntlet:/usr/obj/usr/src/sys/GENERIC amd64
1 [ 0.0%] Tasks: 92, 0 thr; 3 running
2 [ 0.0%] Load average: 0.44 0.55 0.60
3 [ 0.0%] Uptime: 00:33:35
4 [ 0.0%]
Mem[|||||||||||||| 2.95G/15.8G]
Swp[ 0K/8.00G]
PID USER PRI NI VIRT RES S CPU% MEM% TIME+ Command
2509 gjp4 83 0 2722M 2230M R 45.7 13.8 17:49.42 /usr/local/bin/firefox
1716 gjp4 30 10 1494M 979M S 0.0 6.1 1:20.25 /usr/local/bin/sysadm-client
1654 gjp4 20 0 789M 190M S 0.1 1.2 0:40.81 lumina-desktop
1531 root -21 -21 469M 118M S 0.7 0.7 0:29.23 X :0 -auth /var/tmp/.PCDMAuth-0ea04e26e8f074e8ee3e -nolisten tcp
1695 gjp4 30 10 433M 53192 S 0.0 0.3 0:00.72 /usr/local/bin/python2.7 /usr/local/bin/hp-systray -x
1840 gjp4 20 0 400M 49680 S 0.9 0.3 0:12.15 /usr/local/bin/konsole
$
… half-expect leaps to continue. I'll review usage after 22:00 …
Lumina alone. One screenshot three seconds before 22:00, three shots soon afterwards.
22:00:41, a remarkable increase in use of memory:
22:02:32, usage still increasing and the September messages in November are exemplary of #128:
22:03:51, the surge of activity has ceased but the memory footprint of sysadm-client remains much larger than is desirable:
If I had left the computer on with sysadm-client running, then almost certainly I would have found the system effectively useless (swap space exhausted) some time on Sunday.
To avoid that exhaustion I logged out (ending my Lumina test session), logged in to KDE, closed the icon of the secondary SysAdm client (discussion), then used the one remaining SysAdm icon to close the client.
Side note: whilst the closure was effective – there's no longer a sysadm-client process – the icon persists in the System Tray of KDE. This seems to be a rare, negligible, cosmetic bug.
On one hand: no recent incidents of this bug.
On the other hand: I have not recently used PersonaCrypt from a USB flash drive. So I'll not rush to close this issue.
I have a personacrypt usb install. I have not been able to reproduce this issue for some time.
This bugged me on various occasions and I'd like to make it reproducible for you but Murphy's Law, the hardware that I normally use for FreeNAS died a few days ago. On at least one occasion, all attempts to kill the client failed and so it was necessary to force off the computer (press and hold the power button).
I have a hunch that the client becomes hungry in some Life Preserver replication failure situations.
2016-10-18 21:41, hungry sysadm-client
The client as it probably was before the hunger was observed
2016-10-19 05:36, around ten minutes after boot
2016-10-19 05:41, around fifteen minutes after boot
2016-10-19 around 07:51
2016-10-22 14:48, nearly all swap space used
2016-10-22 around 14:50, processes
2016-10-22 around 14:50
2016-10-22 around 14:52, after using System Monitor to kill the client
AFAIR other approaches to killing, probably with htop, did not succeed.
2016-10-22 around 14:57, a view of Life Preserver schedules