Plausible memory leak v1.4.0rc12 at WS-12-250-AC

[Mikea]

Hey all can you confirm the following memory issue.

Device: WS-12-250-AC

F/W: v1.4.0rc12

I ve upgraded to rc12 at April 8 and i m seeing the following mem charts.

Thank you for your time and efforts.

[sirhc]

There was a BIG leak in v1.4.0rc13 which was never released.

Maybe there is a small leak in rc12 I do not know and Eric is off tonight playing poker tonight.

I know he fixed a BAD memory leak in rc13 which filled the memory up in about 6 hours and forced the switch to auto reboot when it reach 100% memory used which is why it was never released.

I will ping him on this so he can look into it and respond tomorrow.

I just looked at my switches running v1.4.0rc12 and they seem OK but they have only been up for 4 days so it is not real noticeable if there is a leak.

But thanks for reporting it.

The next v1.4.0rcX is due out SOON.

[sirhc]

The memory leak was in the Ping Watch dog routine which has been completely re-written since rc12.

Do you have a ping watch dog running?

[Mikea]

from what i can recall no i have not watch dog enabled.
I ll try to look more close on it to see if there are any clues and i ll report back.

[Eric Stern]

Can you run a 'ps' command and post the output?

[Mikea]

Sure Eric, here it is:

Code: Select all

admin@iWiSp_Mountain_Switch:/www# ps
 PID USER VSZ STAT COMMAND
 1 admin 2688 S init
 2 admin 0 SW< [kthreadd]
 3 admin 0 SW< [ksoftirqd/0]
 4 admin 0 SW< [watchdog/0]
 5 admin 0 SW< [events/0]
 6 admin 0 SW< [khelper]
 21 admin 0 SW< [kblockd/0]
 45 admin 0 SW [pdflush]
 46 admin 0 SW [pdflush]
 47 admin 0 SW< [kswapd0]
 48 admin 0 SW< [aio/0]
 58 admin 0 SW< [mtdblockd]
 61 admin 0 SW< [spi-vcoreiii.0]
 269 admin 0 SWN [jffs2_gcd_mtd6]
 284 admin 2688 S init
 369 admin 2676 S klogd
 381 admin 1132 S /sbin/hotplug2 --override --persistent --max-childre
 583 admin 2684 S watchdog -t 500ms /dev/watchdog
 763 admin 2676 S syslogd -S -s128
 1268 admin 19184 S vtss_appl -l -m -s -a -i -p
 1278 admin 19184 S vtss_appl -l -m -s -a -i -p
 1279 admin 19184 S {RX thread} vtss_appl -l -m -s -a -i -p
 1280 admin 19184 S {event thread} vtss_appl -l -m -s -a -i -p
 1298 admin 19184 S {mstp_thread} vtss_appl -l -m -s -a -i -p
 1322 admin 19184 S {lacp_thread} vtss_appl -l -m -s -a -i -p
 1323 admin 19184 S {loop_protection} vtss_appl -l -m -s -a -i -p
 1324 admin 19184 S vtss_appl -l -m -s -a -i -p
 1325 admin 19184 S vtss_appl -l -m -s -a -i -p
 1327 admin 6272 S lighttpd -f /etc/lighttpd.conf
 1328 admin 39144 S /usr/bin/php-fcgi
 1330 admin 39144 S /usr/bin/php-fcgi
 1332 admin 39144 S /usr/bin/php-fcgi
 1334 admin 1452 S /www/stats.fcgi
 1335 admin 19184 S {port thread} vtss_appl -l -m -s -a -i -p
 1336 admin 19184 S {status_thread} vtss_appl -l -m -s -a -i -p
 1337 admin 1452 S /www/stats.fcgi
 1338 admin 1516 S /www/stats.fcgi
 1339 admin 6804 S lighttpd -f /etc/lighttpd.conf
 1340 admin 6892 S lighttpd -f /etc/lighttpd.conf
 1341 admin 6928 S lighttpd -f /etc/lighttpd.conf
 1343 admin 1936 S /usr/sbin/dropbear -p 22 -z 5 -T 60
 1349 admin 39144 S /usr/bin/php-fcgi
 1351 admin 39292 S /usr/bin/php-fcgi
 1352 admin 39144 S /usr/bin/php-fcgi
 1357 nobody 2268 S shellinaboxd -s /shell:admin:root:/root:-/bin/sh --l
 1361 admin 10964 S switch -g
 1363 nobody 2268 S shellinaboxd -s /shell:admin:root:/root:-/bin/sh --l
 1364 admin 10964 S switch -g
 1366 admin 10964 S {discovery threa} switch -g
 1367 admin 10964 S {ubiquiti discov} switch -g
 1368 admin 10964 S {cisco discovery} switch -g
 1370 admin 10964 S {lldp discovery } switch -g
 1371 admin 10964 S {discovery probe} switch -g
 1372 admin 2692 S sh /usr/bin/monitor
 1373 admin 10964 S {i2c thread} switch -g
 1375 admin 10964 S {poe thread} switch -g
 2728 admin 1992 R /usr/sbin/dropbear -p 22 -z 5 -T 60
 2729 admin 2680 S -ash
 2737 admin 2704 S switch -s
 2827 admin 2676 S sleep 10
 2828 admin 2680 S sh -c /bin/ash
 2829 admin 2684 S /bin/ash
 2830 admin 2680 R ps


[Eric Stern]

Looks like the problem is in discovery. I'll fix it.

[Mikea]

I m glad that i helped you to pin point it :)
keep up the great work!

[Mikea]

I can confirm that the latest F/W 1.4.0rc16 fixed the issue.
In the picture attached the memory usage isn't growing after the update.
Thank you Chris & Eric for fast update!