v1.5.5rcX Bug Reports and Comments

[mayheart]

Loaded new firmware onto the switch.

Packet loss returned about 2 minutes after I returned my vlan configuration to the original way it was.

Reverted change and packet loss went away including the vtss_appl messages.

Code: Select all

Oct 2 13:49:43 UI: Configuration changed by root (::ffff:)
Oct 2 13:49:43 UI: VLAN 1 ID: changed from '1' to '997'
Oct 2 13:49:43 UI: VLAN 1 PortSettings: changed from 'EEEEUUUUUUUUTU' to 'EEEEUUUUUUUUTT'
Oct 2 13:49:47 root: removing lan (eth0.997) from firewall zone lan
Oct 2 13:49:51 root: adding lan (eth0.997) to firewall zone lan
Oct 2 13:49:52 system: starting ntpclient
Oct 2 13:50:11 monitor: restarting vtss_appl
Oct 2 13:50:25 monitor: restarting vtss_appl
Oct 2 13:50:38 monitor: restarting vtss_appl
Oct 2 13:50:52 monitor: restarting vtss_appl
Oct 2 13:51:06 monitor: restarting vtss_appl
Oct 2 13:51:19 monitor: restarting vtss_appl
Oct 2 13:51:33 monitor: restarting vtss_appl
Oct 2 13:51:47 monitor: restarting vtss_appl
Oct 2 13:52:00 monitor: restarting vtss_appl
Oct 2 13:52:14 monitor: restarting vtss_appl
Oct 2 13:52:28 monitor: restarting vtss_appl
Oct 2 13:52:41 monitor: restarting vtss_appl


[Stephen]

What was the model of this switch again?

Fortunately the SFP module's you mentioned just arrived in my office and I'm thinking they might actually be contributing a bit too the instability because our result's have diverged.

[mayheart]

I have this plugged into a WS-12-250-AC. I was seeing the same on a WS-26-400-IDC as well.

[Ludvik]

I will see if I can find out what's happening there as well.

Speaking of which, Ludvick - would you mind sharing your "ip addr" output on the switch that you have ipv6 enabled on as well as the switch that you have IP addresses assigned to non-management VLAN's? If necessary feel free to PM me the details. It will help while I am working on some of the other things you mentioned earlier.

Default configuration (ie with IPv6 disabled), only one VLAN created (without IP), connected to network with SLAAC:

Code: Select all


admin@Netonix_Switch:/www# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue
 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
 inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
 inet6 ::1/128 scope host
 valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 1000
 link/ether ec:13:b2:92:67:40 brd ff:ff:ff:ff:ff:ff
 inet 192.168.1.20/24 brd 192.168.1.255 scope global eth0
 inet6 2a00:1919:0:4000:ee13:b2ff:fe92:6740/64 scope global dynamic
 valid_lft 2591958sec preferred_lft 604758sec
 inet6 fe80::ee13:b2ff:fe92:6740/64 scope link
 valid_lft forever preferred_lft forever


admin@Netonix_Switch:/www# ip6tables -L
Chain INPUT (policy ACCEPT)
target prot opt source destination
Chain FORWARD (policy ACCEPT)
target prot opt source destination
Chain OUTPUT (policy ACCEPT)
target prot opt source destination


admin@Netonix_Switch:/www# cat /proc/sys/net/ipv6/conf/eth0/autoconf
1


Adding "2a00:1919::/32" to Access control list doesn't have effect.

If you permit only DHCP or static IPv6 configuration, you must disable autoconf. But I think - for management purposes is SLAAC sufficient and you must respect this. But it must be clearly visible in the configuration, including functional access control.

--
"second problem" with VLAN IP is probably my mistake. Sorry.

[Stephen]

"second problem" with VLAN IP is probably my mistake. Sorry.

No big deal. Honestly that's what tipped me off to this issue with VLAN ip's or changing the default VID of the Management VLAN that I'm trying to fix right now, so it was helpful.

If you permit only DHCP or static IPv6 configuration, you must disable autoconf. But I think - for management purposes is SLAAC sufficient and you must respect this. But it must be clearly visible in the configuration, including functional access control.

So let me make sure I understand what you mean here. Even though IPv6 is disabled, the switch is getting an IPv6 address (i.e. 2a00:1919:0:4000:ee13:b2ff:fe92:6740/64) from SLAAC because auto conf is set to 1:

Code: Select all

admin@Netonix_Switch:/www# cat /proc/sys/net/ipv6/conf/eth0/autoconf
1


But because it is disabled in the interface, you are unable to add this address to the Access Control list thus leaving the switch vulnerable.

Correct?

[Ludvik]

Yes, correct.

[Stephen]

Understood that is definitely important.

[Stephen]

mayheart, I've sent you another firmware to test with. Let me know how it goes.

[mayheart]

Installed the latest one, same issue persists.

Also now I have under voltage warnings, I downgraded it to the older version and it went away. This is a revision F board.

Code: Select all

Jan 1 00:00:10 netonix: 1.5.5rc3-201910040216 on WS-12-250-AC
Jan 1 00:00:13 system: Setting MAC address from flash configuration: EC:13:B2:64:09:0E
Jan 1 00:00:17 root: adding lan (eth0) to firewall zone lan
Dec 31 19:00:33 root: removing lan (eth0.997) from firewall zone lan
Dec 31 19:00:37 root: adding lan (eth0.997) to firewall zone lan
Dec 31 19:00:44 root: adding lan (eth0.997) to firewall zone lan
Dec 31 19:00:46 system: starting ntpclient
Oct 4 07:32:54 dropbear[1362]: Running in background
Oct 4 07:32:56 switch[1389]: temp sensor version 3
Oct 4 07:32:56 switch[1390]: Detected warm boot
Oct 4 07:32:59 monitor: restarting vtss_appl
Oct 4 07:33:11 switch[1381]: 48V is out of range (46.0V)
Oct 4 07:33:13 monitor: restarting vtss_appl
Oct 4 07:33:15 switch[1381]: 24V is out of range (22.8V)
Oct 4 07:33:26 monitor: restarting vtss_appl
Oct 4 07:33:37 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:33:39 monitor: restarting vtss_appl
Oct 4 07:33:53 monitor: restarting vtss_appl
Oct 4 07:34:04 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:34:06 monitor: restarting vtss_appl
Oct 4 07:34:17 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:34:19 monitor: restarting vtss_appl
Oct 4 07:34:33 monitor: restarting vtss_appl
Oct 4 07:34:44 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:34:46 monitor: restarting vtss_appl
Oct 4 07:35:00 monitor: restarting vtss_appl
Oct 4 07:35:10 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:35:13 monitor: restarting vtss_appl
Oct 4 07:35:27 monitor: restarting vtss_appl
Oct 4 07:35:39 switch[1381]: 24V is out of range (22.8V)
Oct 4 07:35:40 monitor: restarting vtss_appl
Oct 4 07:35:52 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:35:54 monitor: restarting vtss_appl
Oct 4 07:35:54 monitor: restarting snmpd
Oct 4 07:36:06 switch[1381]: 24V is out of range (22.8V)
Oct 4 07:36:08 monitor: restarting vtss_appl
Oct 4 07:36:08 monitor: restarting snmpd
Oct 4 07:36:19 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:36:21 monitor: restarting vtss_appl
Oct 4 07:36:21 monitor: restarting snmpd
Oct 4 07:36:35 monitor: restarting vtss_appl
Oct 4 07:36:47 switch[1381]: 24V is out of range (22.8V)
Oct 4 07:36:48 monitor: restarting vtss_appl
Oct 4 07:36:49 monitor: restarting snmpd
Oct 4 07:37:00 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:37:02 monitor: restarting vtss_appl
Oct 4 07:37:02 monitor: restarting snmpd
Oct 4 07:37:16 monitor: restarting vtss_appl
Oct 4 07:37:16 monitor: restarting snmpd
Oct 4 07:37:26 switch[1381]: 24V is out of range (22.5V)
Oct 4 07:37:30 monitor: restarting vtss_appl
Oct 4 07:37:43 monitor: restarting vtss_appl
Oct 4 07:37:43 monitor: restarting snmpd


[sirhc]

Also now I have under voltage warnings, I downgraded it to the older version and it went away. This is a revision F board.

So there was a recent "small" batch, less than 200, of WS-12-250-AC units that went out with the wrong voltage trim. This was a mess up from China factory that makes those for us. They trimmed them too low and we did not catch it right away as it was right on the edge and our tech who serializes for us only noticed those that failed and those that passed right on the edge went through. Was not until someone had mentioned to me that I looked into it and discovered the issues. We had to manually trim all the power supplies from that shipment, some 2,000 power supplies.....YEA EXTRA WORK. But sadly 200+/- had already shipped with lower than normal trims.

It is simple to fix if you want. Open the chassis and near the DC output side of the AC power supply is a SQUARE BLUE TRIM POT with a small Phillips adjustment that is sealed with green stuff. Just use a small crew driver and gently turn it while monitoring the Device Status TAB 48V. We normally want that at 50V. You can also use a volt ohm meter connected to power supply output and you would want to trim it there at 51V.

Or you can RMA it and we fix for free.