v2.0.8rc1 Bug Reports and Comments - WS3 Firmware

[sirhc]

Release Candidate v2.0.8rc1

Click HERE to download firmware v2.0.8rc1 - Released 8/16/2022
Click HERE to go to the thread for v2.0.8rc1 Bug Reports and Comments.

FIXED/CHANGED
- Issues with SFP+ configuration
- Memory leak that caused reboots
- Security Issues
- SNMP Issues

ENHANCEMENTS

KNOWN ISSUES
- WEB UI issues when not at 100% Zoom on browser especially on VLAN TAB
- Some language templates need help

Released 8/16/2022

[mayheart]

So far it's working well.

I did have one issue with one switch running 2.0.7 that required a power cycle to bring the unit back online after sending out an upgrade.

The external temperature/humidity sensor, is there any OIDs for the value it connects that can be graphed via SNMP?

[mayheart]

In the last 2 hours or so, cpu usage has been stuck at 100%

STP, Loop protection and all discovery options are turned off. Only SNMP is enabled.

Code: Select all

Mem: 123236K used, 129952K free, 0K shrd, 25148K buff, 42184K cached
CPU: 69% usr 29% sys 0% nic 0% idle 0% io 0% irq 1% sirq
Load average: 2.72 2.65 2.45 3/94 7994
 PID PPID USER STAT VSZ %VSZ %CPU COMMAND
 45 1 admin S 476m 193% 51% /usr/bin/switch_app
 602 1 admin R 2448 1% 42% netonix_app -c
 7928 7919 admin R 3260 1% 2% top
 7 2 admin SW 0 0% 1% [ksoftirqd/0]
 7914 208 admin S 5428 2% 1% dropbear -z 5 -T 60
 19 2 admin IW 0 0% 0% [kworker/0:1-eve]
 145 143 admin S 140m 57% 0% php-fpm: pool www
 146 143 admin S 140m 57% 0% php-fpm: pool www
 144 143 admin S 140m 57% 0% php-fpm: pool www
 143 1 admin S 140m 57% 0% php-fpm: master process (/tmp/php-fpm.conf)
 218 1 admin S 10580 4% 0% ntpd -g 192.168.94.137
 179 1 admin S 10576 4% 0% snmpd -Lf /dev/null -c /tmp/config/snmpd -I system_mib dlmod memory loadave disk va
 226 224 admin S 9736 4% 0% lighttpd -f /tmp/lighttpd.conf
 227 224 admin S 9732 4% 0% lighttpd -f /tmp/lighttpd.conf
 225 224 admin S 9732 4% 0% lighttpd -f /tmp/lighttpd.conf
 224 1 admin S 9192 4% 0% lighttpd -f /tmp/lighttpd.conf
 216 214 nobody S 8548 3% 0% shellinaboxd -t -s /shell:admin:root:/root:/bin/sh -l --localhost-only -a TJd8LMgec
 214 1 nobody S 8432 3% 0% shellinaboxd -t -s /shell:admin:root:/root:/bin/sh -l --localhost-only -a TJd8LMgec
 92 45 admin S 7764 3% 0% /usr/sbin/zebra -f /etc/quagga/zebra.conf -i /tmp/zebra.pid -P 0 -z /tmp/zebra.sock
 228 1 admin S 5768 2% 0% {monitor} /bin/sh /usr/bin/monitor
 7994 228 admin S 5768 2% 0% sleep 10
 153 1 admin S 5768 2% 0% syslogd -S -s128
 208 1 admin S 5292 2% 0% dropbear -z 5 -T 60
 7919 7918 admin S 3260 1% 0% /bin/sh
 7915 7914 admin S 3124 1% 0% -sh
 7918 7915 admin S 2448 1% 0% netonix_app -c
 1 0 admin S 1500 1% 0% /usr/bin/stage2-loader
 42 2 admin IW< 0 0% 0% [kworker/0:1H-kb]
 41 2 admin SW< 0 0% 0% [loop0]
 36 2 admin SW 0 0% 0% [spi0]
 39 2 admin SW 0 0% 0% [ubi_bgt0d]
 43 2 admin SW< 0 0% 0% [loop1]
 5 2 admin IW 0 0% 0% [kworker/u2:0-ev]
 8 2 admin SW 0 0% 0% [kdevtmpfs]
 44 2 admin SW< 0 0% 0% [loop2]
 2 0 admin SW 0 0% 0% [kthreadd]
 4 2 admin IW< 0 0% 0% [kworker/0:0H-kb]
 6 2 admin IW< 0 0% 0% [mm_percpu_wq]
 9 2 admin SW 0 0% 0% [oom_reaper]
 10 2 admin IW< 0 0% 0% [writeback]
 11 2 admin SW 0 0% 0% [kcompactd0]
 12 2 admin IW 0 0% 0% [kworker/u2:1-ev]
 34 2 admin IW< 0 0% 0% [kblockd]
 35 2 admin SW 0 0% 0% [kswapd0]
 37 2 admin IW< 0 0% 0% [ipv6_addrconf]
 38 2 admin IW 0 0% 0% [kworker/0:2]


[wtm]

Well, loaded the 2.0.8 firmware on several of the WS-3 units, and it seemed to take it OK.

HOWEVER, we then went to program the configurations on several, and that is where we run into problems, and the units are un-usable again !

Programming seems to be glitchy. If you change a configuration parameter that effects the browser url, then it normally comes up with a popup that it will reload in 5 seconds. Sometimes that happens and sometimes it doesn't. Other parameter changes normally have the browser pop back up after about 30 seconds when those changes are done. But not in these WS-3 units. You have to wait for 30 seconds or more to go by, then physically reload your browser.

Both Firefox and Chrome was used.

Then for no reason, you will change another parameter, such as the password, or un-check the "Flow Control" on the Storm Broadcast. THen the browser will go into it's countdown, and you can NEVER get back into the switch. The IP address that was programmed into it is no longer available at that point.

Also, you can occassionally program a configuration, such as the password, and it will NOT take it, although it will show that it has on the configuration page. The old password still needs to be used on the login.
Not sure what needs to be looked at, but if you can't program it, and it won't take it, you can't use it.

[mayheart]

I had a power blip at this site.

After about 3 hours of uptime the switch went back to 100% cpu.

[mayheart]

Any update on this? The switches webui become pretty useless when its at max cpu.

[wtm]

I think they have basically given up on doing anything on the WS-3 units ! Sad........

[Dave]

no we haven't given up...we have a lot of them being help up in production waiting for code...

[michaeln416]

Is there an estimate on when 2.0.8rc1 may become fully baked and upgraded to a stable release?

I'm still running 2.0.7 in production and getting spontaneous reboots every 30 - 60 days.

We are hesitant to log into the Web UI to make configuration changes as that increases the risk of a spontaneous reboot or locking things up.

This is painful.

Mike