Westermo RN Weos v4 32 3
Westermo RN Weos v4 32 3
Contents
1 Summary of Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1 News in WeOS 4.32.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.1 Modbus Firewall (DPI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.2 Ability to Disable HTTP and HTTPS Services Individually . . . . . . . . . . . . 6
1.1.3 ’ipconfig’ Service Limited to Discovery and IP Settings . . . . . . . . . . . . . . 7
1.1.4 New Default Boot-loader Version . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.5 Fixed CVEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.6 Web/HTTPS Support for TLS v1.0 and v1.1 Removed . . . . . . . . . . . . . . 8
1.2 News in WeOS 4.32.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 News in WeOS 4.32.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.1 Fixed CVEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 News in WeOS 4.32.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4.1 New bootloader with fixes for ’Corazon’ RFI/RFIR products . . . . . . . . . . . 8
2 Known Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1 Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 CLI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 NAT translation on ’Basis’ products . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 DHCP Static Lease Preemption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.5 DHCP classless static routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.6 IEEE 802.1X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.7 MRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.8 SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.9 Web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.10 IPsec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.11 SSL VPN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.12 Link Aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.13 Bandwidth Limiting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.14 Flow Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.15 LLDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.16 TTDP (IEC 61375-2-5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3 Fixed Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.1 WeOS 4.32.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 WeOS 4.32.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.3 WeOS 4.32.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4 WeOS 4.32.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4 Known Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5 Technology Previews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.1 Password Encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2 Interface Admin Distance Trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3 Interface Gateway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.4 SFP DDM Alarm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.5 Serial Low Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.6 SNMP RIPv2-MIB and OSPF-MIB . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.7 NTP Server with GPS Time-Base Support . . . . . . . . . . . . . . . . . . . . . . . 24
5.8 TFTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.9 Preferred (Remote) NTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.10 Guest User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.11 New VLAN Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.12 Additional SSH Server Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.13 Additional Telnet Server Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.14 Packet Capture – TCP Dump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.15 CLI Welcome Message . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.16 Additional DDNS Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.17 USB Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.18 IPsec and SSL VPN Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.19 PPPoE Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.20 Serial HDX Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.21 DHCP Client ”ARP Ping” Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.22 Support for Disabling DHCP Snooping in DHCP Relay Agent . . . . . . . . . . . . . 30
5.23 Firewall Conntrack Flushing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.24 RSTP Support for VLAN Tagging of BPDUs . . . . . . . . . . . . . . . . . . . . . . 31
5.25 Remote IO Support for Digital Output . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.26 Storm Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6 Accessing the Command Line Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7 Firmware Upgrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.1 What Firmware Image to Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.2 Upgrading the Primary Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.3 Upgrading the Boot Loader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7.4 Special Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7.4.1 Upgrading ’Basis’ products with version lower than 4.29.0 . . . . . . . . . . . . 37
7.4.2 Upgrading Viper 12A and 20A with version lower than 4.22.0 . . . . . . . . . . 39
7.4.3 Upgrading units with lower version than 4.13.1 . . . . . . . . . . . . . . . . . . 40
Legal Information
The contents of this document are provided “as is”. Except as required by applicable law, no war-
ranties of any kind, either express or implied, including, but not limited to, the implied warranties
of merchantability and fitness for a particular purpose, are made in relation to the accuracy and re-
liability or contents of this document. Westermo reserves the right to revise this document or with-
draw it at any time without prior notice. Under no circumstances shall Westermo be responsible for
any loss of data or income or any special, incidental, and consequential or indirect damages how-
soever caused. More information about Westermo can be found at the following Internet address:
http://www.westermo.com.
About
Westermo WeOS is a network operating system designed for industrial grade rugged Ethernet switches
and routers. Fully supporting WeOS units on Basis platform (Lynx 2nd gen., Wolverine 2nd gen., Viper
2nd gen., and Falcon), Corazon platform (RedFox 2nd gen.) and Coronet platform (Viper-A, a.k.a,
Viper 3rd gen.).
WeOS is a Linux based software platform that has been in operation since 2006 on custom made
RedFox Mil, RedFox Aero and RedFox Rail products. With the advent of the RedFox Industrial
line of products the platform was given a major overhaul to improve standards compliance as well
as compatibility requirements with existing Westermo product offerings. The result is WeOS, the
Westermo Operating System.
For more information about Westermo and other product offerings see http://www.westermo.
com/.
1 Summary of Changes
1.1 News in WeOS 4.32.0
WeOS 4.32.0 contains new or updated features as described in the subsections below. It also includes
bug fixes as listed in section 3.1.
web
no port
ssl-port 443
end
example:/#>
Note: ’no port’ and ’no ssl-port’ commands both have new semantics in 4.32.0 and onward. Now they
mean ’disable’ HTTP and HTTPS respectively, while in earlier releases they implied reset to default
port (’port 80’ and ’ssl-port 443’ respectively).
See the ’General System Settings’ chapter of the WeOS Management Guide for information on how
to manage the Web service.
See the ’General System Settings’ chapter of the WeOS Management Guide for information on how
to manage the ipconfig service.
This new Barebox version contains no changes applicable for WeOS 4 units as compared to the previous
versions included in the WeOS 4.29, 4.30 and 4.31 releases. The reason for including this new version
is that Westermo’s ambition is to have a common preferred Barebox version both for WeOS 4 and
WeOS 5.
In general, there is no need to upgrade WeOS units in the field to the latest boot-loader, however, see
section 7.4 for special considerations.
• CVE-2018-12684
• CVE-2020-27304
• CVE-2022-0778
2 Known Limitations
This section includes known reported bugs and missing features, which may not necessarily be limi-
tations, in many cases they may constitute severe operational drawbacks.
2.1 Platform
• A system with many VLANs setup requires more time at boot. This was first reported in #3291,
but even after having fully optimised all data paths there still remains a significant delay. E.g.,
creating 128 VLANs on a RedFox Industrial takes apx. 6 seconds longer than creating a single
VLAN.
• Running an FRNT ring over copper SFPs is not recommended, due to slow response time from
copper SFPs.
• Limited support for low–level interaction with PHYs and link partners.
• When toggling bridge priority on the elected root bridge storm is easily provoked, issue #4203.
Avoid reconfiguring bridge priority at runtime.
• In some setups when RSTP gets link up it has been reported to take very long to reconfigure,
issue #4707.
2.2 CLI
• When issuing, e.g., show running not all settings are shown. This is due to WeOS 4.3.0 and
later only showing differences to the system default. Support for show running [all],
where the optional ’all’ keyword would list everything, is planned for a later release.
• The on–line help is not only insufficient, it is sometimes even misleading. E.g., some commands
do not support the no prefix, some commands do not support show and no commands in
configure context support repeat. Cleanup and improvement is a work in progress.
For information about products and which platform they are based on, see the ’Introduction’ Chapter
in the WeOS Management Guide.
This can be achieved by carefully choosing which routes are put in the global, subnet and
host scopes, and by consolidating several smaller routable CIDR networks to fewer larger
ones.
This issue starts to become relevant at around 15 routes sent to a single client, if the minimum
legal DHCP message size is requested, and if options 121 and 249 are used simultaneously. Due
to certain peculiarities in how the DHCP options are defined, this limit is somewhat dependent
on the routes in question.
• DHCP server limitation
When a high number of routes is configured for an individual scope - including routes inherited
from other scopes - the DHCP server may be unable to correctly function, leading to a loss of
DHCP service. This limit is at about - again, depending on the specific values used - 33 routes.
This issue can be mitigated by performing the following action:
– Ensure that any individual lease will only contain the routes it needs.
This can be achieved by carefully choosing which routes are put in the global, subnet and
host scopes, and by consolidating several smaller routable CIDR networks to fewer larger
ones.
2.7 MRP
MRP ring ports have to be on the same slot when used on products with multiple slots. MRP does not
support aggregate ports.
2.8 SNMP
The SNMP chapter of the WeOS Management Guide lists supported standard MIBs, including limita-
tions to specific tables for some MIBs. Additional deviations from the standard MIBs may exist. For
some MIBs, you find more detailed MIB conformance information in the WeOS release zip-archive.
2.9 Web
• Inspecting RMON counters in the Port Statistics page may need a manual reload before the
actual values are displayed.
• Due to security reasons the username and password must always be provided when logging in,
i.e auto-completion is not supported in the login form.
2.10 IPsec
• MTU override may not work as expected, sending a message over the IPsec tunnel will not
respect mtu override on the other end. Workaround: Always have the same MTU on the
interfaces on both ends of the tunnel.
• The remote IP address of the IPsec gateway may in some circumstances not be reachable from
an IP address associated with the IPsec tunnel.
Workaround: Always connect to an IP on the IPsec gateway that is reachable from within the
tunnel.
• DPD restart/clear is sometimes unreliable. If the responder is configured with dpd-action clear
and then rebooted, the tunnel will sometimes not be renegotiated.
Workaround 1: If only using static IP addresses and only one initiator, change both nodes to
be initiators and set dpd-action hold on both sites.
Workaround 2: Use a ping trigger towards an IP inside the tunnel and connect the ping trigger to
a tunnel action (See section 5.18). Remember to set ’retrain interval 30’ in the action configured,
this will create another level of DPD outside the IKE traffic, but it will be encrypted.
Workaround 3: If it is not possible to use DPD hold (multiple initiators or not static IP) you
can on the initiator(s), create a tunnel action (see section 5.18) and set it to be retrained after a
few seconds. Use a ping trigger and set peer as an IP inside the tunnel, and connect it to the
tunnel action.
• To be able to use dynamic or static routing over a ssl interface you will have to use a layer 2
tunnel. Layer 3 tunnels will not work as expected in this scenario.
• No check for certificate type, a client certificate can be used as a server certificate and reversed.
• When using layer3, OpenVPN supports multiple topologies, p2p, net30 and subnet, WeOS only
support subnet.
• VLAN support: There is no support to add a link aggregate to a VLAN. Instead, each of the
individual member links need to be added to the appropriate VLANs.
• Port settings: There is no support to configure port settings for the link aggregate. Instead,
each of the individual member ports need to be configured uniformly, e.g., with respect to port
speed/duplex mode.
• Only link aggregation of Ethernet ports is supported. Configuration of SHDSL ports or xDSL
ports (ADSL/VDSL) ports in an aggregate, or mixing Ethernet and SHDSL/xDSL ports in an
aggregate, may be possible, but this is not supported and the behaviour is undefined.
• not supported on certain RedFox models. See management guide for further details.
Note: There are currently no warnings when the ”fps” does not apply for traffic shaping. In such
cases, the rate setting will ignore the fps attribute and interpret the rate in bps.
• When enabling 802.3x flow-control on a port, that port will permanently operate in flow-control
mode without involving Ethernet auto-negotiation mechanisms.
• When configuring flow-control in ”slot based” WeOS products (for example RFI), a port with
802.3x flow-control enabled will only send Pause frames when causing congestion on another
port in the same slot; not when causing congestion on a port in another slot. For example, if port
1/1 has flow control enabled, it will send Pause frames when causing congestion on on port 1/2,
but not when causing congestion on port 2/1. Similar restrictions apply to WeOS Viper, RedFox
Rail (RFR) and RedFox Industrial Rack (RFIR) products.
2.15 LLDP
The WeOS LLDP support has the following known limitation:
When connected to Windows 10 PC with LLDP driver enabled, the WeOS unit will not respond to
SNMP requests for the “remsysname” LLDP SNMP OID. The proposed work-around is to disable the
LLDP driver on the Windows 10 PC.
3 Fixed Issues
3.1 WeOS 4.32.0
Fixed issues in WeOS 4.32.0 (as relative to 4.31.0).
Note: Issue #14307 was fixed already in WeOS 4.29.0, but was not included in the list of fixed issues
until release notes of 4.32.0.
Note: Issue #18264 was fixed in WeOS 4.29.0, but was not included in the list of fixed issues until
release notes of 4.32.1.
WeOS 4.32.3 comes with a new Barebox boot loader version. Updating the boot loader to this version
fixes an issue on some ’Corazon’ RFI/RFIR products as stated in section 1.4.1.
4 Known Issues
Known issues as of WeOS 4.32.3.
The following three issues have been removed from the list of known issues in 4.32.1 and 4.32.2, and
are instead listed as known limitations, see sections 2.4 (#18353), 2.6 (#17295), and 2.3 (#14378)
respectively.
5 Technology Previews
WeOS contains hidden and undocumented features called technology previews. Westermo provides no
support for undocumented features. Features specifically marked as tech previews can be completely
redesigned, removed or changed in such a way that after an upgrade they are not guaranteed to work!
The following is by no means a complete list, but details features that may become supported in the
next upcoming feature release.
• Password Encryption: CLI only, section 5.1
• Trigger to control Interface Admin Distance: CLI only, section 5.2
• Setting default gateway per interface: CLI only, section 5.3)
• SFP DDM Alarm and SNMP Trap Support: Alarm settings in CLI only, section 5.4
• Serial Low Latency: CLI only, section 5.5
• SNMP MIBs for RIP and OSPF: See section 5.6
• NTP Server with GPS support: CLI only, section 5.7
• TFTP Server: CLI only, section 5.8
• Guest User: CLI only, section 5.10
• New VLAN Features: CLI only support for disabling ’secure’ mode and MAC address learning.
See section 5.11
• Additional SSH settings: CLI only support for setting SSL port, idle-timeout and keepalive
interval for the WeOS SSH server. See section 5.12
• Additional Telnet settings: CLI only support for setting Telnet port for the WeOS Telnet server.
See section 5.13
• Tcpdump: CLI only. See section 5.14
• CLI welcome message: Ability to set custom CLI welcome message. See section 5.15
• Additional DDNS features: See section 5.16
• USB boot: CLI only, section 5.17.
Separate feature from ”USB Autobackup/restore” and ”USB Configuration Deployment”!
• IPsec and SSL VPN extensions: Ability for IPsec initiators to be configured with two respon-
der addresses (IPsec Backup Peer), IPsec and SSL VPN tunnels can be enabled/disabled via
alarm trigger and action, and SSL VPN tunnels can be configured using a standard OpenVPN
configuration file (.ovpn). See section 5.18.
To further secure an installation the user can provide a custom encryption key. This key will be device
specific and must be entered again if exporting the configuration to another device. The key can be at
most 64 characters long and will be securely1 stored in built-in flash of the device to be able to boot.
example:/#> config
example:/config/#> encrypt passwords key XYZZY
example:/config/#> leave
example:/#> copy run start
To change custom key from ’XYZZY’ to ’QWERTY’ the user will be prompted to input the current
custom encryption key. This prompt will not appear when changing from the default built-in key. To
change from a custom key back to the default built-in key type:
example:/#> config
example:/config/#> encrypt passwords default
1 The custom key is in itself encrypted before stored in a file on built-in flash.
Password encryption is a per-file feature of WeOS. If you change to another configuration file, using the
copy command, that file determines if password encryption is enabled or disabled. When changing
to a file with encryption you can be prompted for its secret key, if a custom secret key was used to
encrypt its passwords.
When disabling password encryption, using the no encrypt command, all password strings will be
scrambled using a random secret key. This maybe seems a bit unintuitive, but is a security measure
to protect your secrets from being decrypted by someone with access to a copy of your encrypted
configuration and rogue WeOS device.
A factory reset, using crossed cables or the factory reset login on the console, will wipe all configura-
tions, including any custom secret keys.
NOTE: Make sure to configure the trigger (ID 2 in this example) to use the correct outbound interface,
otherwise the ping will use the default route, and you will get interesting flapping.
Please note, this setting is only available for interfaces of type inet static.
For ’rx-power’, the alarm condition should be set to ’low’ as a low value means a weaker signal. For
values below the falling threshold, alarm will be indicated.
example:/config/#> alarm
example:/config/alarm/#> trigger ddm-rx-power (Create DDM temp. trigger)
example:/config/alarm/trigger-4/#> condition low (Alarm on low signal)
example:/config/alarm/trigger-4/#> port 1 (Trigger on SFP port 1)
example:/config/alarm/trigger-4/#> threshold rising -20 (Rising threshold at -20 dBm)
example:/config/alarm/trigger-4/#> threshold falling -25 (Falling threshold at -25 dBm)
Hint: To limit access to the NTP server to a specific interface you currently have to use the WeOS
firewall functionality (firewall functionality is available on the 200-series of WeOS products).
Technology preview additions to NTP server function:
As a complementary technology preview, it is possible to let the WeOS unit be the top-level NTP
server, using it’s own system clock as source to synchronise time on connected NTP clients.
WeOS also have support for using a GPS receiver connected using RS232/422/485 and use it as a
reference clock for NTP. This requires a preconfigured GPS receiver, it has to be configured to send
NMEA reports. The correct serial port configuration also has to be entered into the serial port context
in the CLI. Pulse per second (PPS) is currently not supported, but may be supported in future releases.
In the example above, the GPS was attached to serial port ”1”. Additional configuration of serial port 1
(e.g., bit-rate) may be required to match your GPS.
The guest account is very restricted, e.g., it cannot configure the system, read passwords by from
configuration files, or otherwise manipulate the state of the system. Only inspect status of ports,
VLANs, interfaces and RMON, and do basic network debugging using, e.g., ping or traceroute.
The exposed tcpdump features are limited, but should be sufficient for most use-cases. One such
feature is the ability to save the PCAP files to a USB stick, if the device is equipped with a USB port.
See the online help in the CLI for more information and some useful examples to get started.
• Console login:
example:/config/system/#> [no] issue <MESSAGE>
• SSH login:
example:/config/system/#> [no] issue-net <MESSAGE>
Example:
The messages of the ’issue’ and ’issue-net’ settings are limited to 64 characters, while the ’motd’
message can be longer.
The WeOS DDNS client, Inadyn, now has support for a few more DDNS providers: 3322, ZoneEdit,
easyDNS, DNS-O-Matic, ChangeIP, nsupdate.info, DuckDNS, and Loopia. This in addition to the
already supported: DynDNS, FreeDNS, and No-IP.
HTTPS/SSL Support
Some DDNS providers support HTTPS update, this WeOS 4.15.1 and later support an SSL check box
in the WebUI and a ’ssl’ setting in the CLI to enable this feature. Please note, you need to make sure
your DDNS provider supports this before enabling SSL.
Forced Update
The DDNS client only sends an update to your DDNS provider when the IP address changes, and
a forced update every week. In some cases, however, you may need to manually force an update.
Currently this is only possible from the CLI (web support is planned for a later release)
See the system log file for both the action and results of the update. The actual DNS update may
take a while to propagate to your Internet Service Provider (ISP), so please don’t issue this command
multiple times thinking this will speed up the process. It all depends on how your DNS record is setup
at the DDNS provider.
Also, if you do this too often some DDNS providers will disable your account, or your DNS entry, for
excessive updates. This is a policy of the DDNS provider.
See the WeOS Management Guide (in particular the chapter ”General Switch Maintenance”) for more
information on USB support (bootstrap settings, supported USB sticks and file systems, etc.)
• IPsec Backup Peer This is a technology preview of upcoming IPsec redundancy support.
IPsec initiators may be configured with two responder addresses. If IPsec fails to connect to the
primary responder, it will try to connect to the backup responder. The primary responder will
periodically be checked, and a switch back is initiated if possible.
example:/config/alarm/#> trigger ping
example:/config/alarm/trigger-1/#> peer 192.168.22.2
example:/config/alarm/trigger-1/#> end
example:/config/alarm/#> end
example:/config/#> tunnel
example:/config/tunnel/ipsec-0/#> backup 192.168.23.2 trigger 1
example:/config/alarm/#> leave
• Enable/disable IPsec and SSL VPN tunnels via alarm trigger and action
example:/config/alarm/action-4/#> target tunnel
example:/config/alarm/action-4/#> tunnel ipsec 0
This can be used to for example have a service tunnel that you want to enable from digital in.
In that case you just create a digital in trigger and connect it to the action created above. When
the trigger is “true”, the VPN tunnel will be enabled.
It is also possible to retrain the tunnel, to not keep it in error state, this is useful if you want to
restart the tunnel on an event:
example:/config/alarm/action-4/#> retrain interval 30
This function is limited to handle RTS-CTS delay of 23 ms and a guard time for the DCD signal of 10
ms. By default, serial HDX mode is disabled.
If the arping setting is disabled, the DHCP client in WeOS sets the IP address assigned by the DHCP
server without first performing an ARP ping of the new IP Address.
By default, the arping option is enabled.
The problem with disabling DHCP snooping is that in “flat” networks where the DHCP client, relay
and server are in the same broadcast domain (same LAN). The DHCP server will receive two DHCP
requests from the client. The recommended workaround is to run the DHCP server on a different port,
e.g. 6767, and have the relay agent forward all requests to the server on that port. That way the server
will ignore broadcasted DHCP requests. However, this requires that all client requests pass through a
relay agent, which in many setups may not be possible.
The connection tracking mechanism is an optimisation in the firewall. Firewall rules are only evaluated
once per connection, and are placed in a cache. This cache speeds things up for the rest of the packets
belonging to the same session.
This may have some side effects if dynamic routing is enabled. A deny rule on a specific interface
may not be respected if a connection is enabled through some other interface and then moves to the
interface through dynamic routing events.
Enabling automatic flushing on route events makes traffic to be re-evaluated in the firewall at route
changes, thus solving this problem.
Note: NAT also uses the same connection tracking cache for its internal state. Flushing the cache may
result in that existing NAT:ed connections can break and reset. Please use with care!
Since WeOS 4.14.2, this setting flushes everything in the connection tracking cache at routing events.
This feature will be changed in a future version to enable a more selective flushing that will avoid
flushing connections that are not affected by a specific routing change.
This feature does not affect reception of RSTP BPDUs, nor does it introduce support for RSTP per
VLAN, or any similar variant offered by other manufacturers. All it does is on a per-port basis enable
a WeOS device to add an IEEE 802.1Q VLAN tag to all BPDUs egressing an RSTP port.
At this point it is unclear if this feature will ever become anything more than a technology preview.
Instead, later versions of WeOS are more likely to add actual RSTP per VLAN support, or even true
MSTP.
Alarm action 1 is used unless an other action is set. The standard configuration action 1 sets the digital
out as one of its targets.
The digital output is active for a certain time. For making the output active for 25000 milliseconds (25
s) request this URL:
http://192.168.2.200/cgi-bin/adm/io?timer1=25000
example:/config/#> alarm
example:/config/alarm/#> action 2
example:/config/alarm/action-2/#> target port led
example:/config/alarm/action-2/#> end
example:/config/alarm/#> trigger storm-detect
example:/config/alarm/trigger-2/#> action 2
example:/config/alarm/trigger-2/#> port 1-2
example:/config/alarm/trigger-2/#> persistent 300
example:/config/alarm/trigger-2/#> threshold 500
example:/config/alarm/trigger-2/#> end
example:/config/alarm/#> show
Trigger Type Enabled Action Source
==========================================================================
1 frnt YES 1 Instance 1
2 storm-detect YES 2 1-2
Action Targets
==========================================================================
1 snmp log led digout
2 log led port
Recommended Clients
UNIX OpenSSH, http://www.openssh.com
Win32 PuTTY, http://www.chiark.greenend.org.uk/~sgtatham/putty/, note that
PuTTY is also useful for connecting to serial port consoles.
Please follow the directions for installation and usage applicable to your system and client.
Logging In
To gain access to the CLI you need:
• An SSH client
• The switch IP#
• The user name and password
Units shipping with WeOS have by default all ports assigned untagged to VLAN 13 , and is configured
to acquire an IP address via DHCP, but also with a static IP address: 192.168.2.200 with netmask
255.255.255.0. The unit’s will also be reachable via a link-local address, i.e, an address in range
169.254.x.x (where ’x’ is a number 0-255).
Use the WeConfig tool, an LLDP client or nmap to find your device. If you have a DHCP server
available you can set it up to hand out a known IP addresses for the registered devices MAC addresses.
Each unit comes with 16 or 32 MAC addresses assigned, depending on the port count, the base address
should be printed on the box and on the unit itself.
The unit is fairly quick to boot, in under 10 seconds is the unit up requesting an IP address — depending
on the existence of a DHCP server the fall back to link–local address can take a while. To be on the
safe side while scanning for your device, expect it to take anything from 30 seconds to one minute
after power–on.
3 Falcon units come with a slightly different factory configuration. The Ethernet ports on Falcon belong to VLAN1 and
are reachable via IP address 192.168.2.200. The xDSL port belongs to VLAN1006 and use DHCP for address assignment.
The following example illustrates how to login to the switch using OpenSSH from a GNU/Linux based
host system. The process is similar with PuTTY or other SSH clients.
The operator lists the running configuration with the command show running, an overview of
ports, vlans and interfaces is available by typing show ports, show vlans and show ifaces.
See the help or the show tutorial for more on line help.
To change some settings, enter the configuration context with the command conf, short for “config-
ure”. The same commands as shown above also apply here, but now display configured settings.
To show or change the interface and VLAN properties the operator uses the command: interface
vlan2 and vlan 2, respectively, with an optional “show” as prefix. E.g. show iface vlan2.
To leave a level the operator must use the command end to save and abort to cancel.
Any new settings are activated only when the operator leaves the configuration context, using “end”.
To save settings to non–volatile RAM (flash disk), the operation uses copy run start from admin–
exec context.
$ ping -b 192.168.2.255
PING 192.168.2.255 (192.168.2.255) 56(84) bytes of data.
64 bytes from 192.168.2.200: icmp_seq=1 ttl=64 time=10.4 ms
64 bytes from 192.168.2.200: icmp_seq=2 ttl=64 time=0.895 ms
ˆC
$ ssh admin@192.168.2.200
The authenticity of host '192.168.2.200 (192.168.2.200)' can't be established.
RSA key fingerprint is 1d:ce:fe:4b:8e:c2:73:42:11:68:73:02:e5:a6:e4:8b.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '192.168.2.200' (RSA) to the list of known hosts.
admin@192.168.2.200's password: westermo
.--.--.--.-----.-----.------.-----.-.--.--------.-----.
| | | | -__|__ --|_ _| -__| _| . . | _ | http://www.westermo.com
\__/\__/|_____|_____| |__| |_____|__| |__|__|__|_____| info@westermo.se
Robust Industrial Data Communications -- Made Easy
This is a typical session where broadcast ping is first used to locate the device, and then SSH login
using the default user and password. (Hint: Use the WeConfig tool to locate your WeOS devices.)
7 Firmware Upgrade
Firmware upgrade is supported from the CLI, Web and WeConfig tool. All of them support FTP/TFTP
upgrade, but the Web also supports CGI upload from the browser – making it the ultimate choice if
you have no FTP/TFTP server available or do not care to set one up.
The version string listed in the output from the show system-information command is only
updated after reboot.
Done.
The system will force a reboot when upgrading the primary image. This to protect against flash cor-
ruption issues seen in earlier releases, caused by simultaneous access to the flash during programming
or when starting new processes after upgrade.
As usual, when upgrading from an earlier release, we always recommend saving your startup config-
uration beforehand.
needs to be upgraded. If the boot loader version is “4.11” or similar, it is a “Redboot” boot
loader and also needs to be upgraded as described below.
• Upgrade the boot loader of your Basis unit.
When upgrading the boot loader, it is important that the unit does not lose power
during the (short) flashing phase.
The new boot loader firmware is included in WeOS-4.32.3.pkg. The bootloader can be
upgraded via Web or WeConfig. In CLI it could look like (assuming FTP/TFTP server at
192.168.2.3):
example:/#> upgrade boot 192.168.2.3 WeOS-4.32.3.pkg
• When the boot loader upgrade is done, reboot the unit.
5. Upgrade to WeOS-4.32.3
Now you can continue upgrading as usual. An example of upgrading the primary image via the
CLI is shown below.
example:/#> upgrade primary 192.168.2.3 WeOS-4.32.3.pkg
The unit will reboot when the upgrade has finished. (It is also recommended to upgrade the
secondary image, but this is not shown here).
This is was the end of the procedure.
A final note on upgrade order. In the description above, the recommendation is to upgrade the primary
image before you upgrade the secondary image:
• If you anyway have upgraded the secondary image to WeOS 4.32.3 on your Basis unit first,
you may experience that this secondary image version is shown as ’0.00’ when listing system
information. This happens when the unit is running a primary image older than 4.29.0 (say
4.28.6), as it is then unable to correctly read the version number for images with the new
(SquashFS) file format.
• Showing the secondary image version number as ’0.00’ is normal in this situation. It will be
shown correctly the once primary image has been upgraded to 4.29.0 or higher. Still, you may
wish to verify that the secondary image was installed correctly before upgrading the primary. If
you have console access to your unit and Barebox installed, you can boot the secondary image
via the Bootmenu. See the WeOS Management Guide (chapter ’General Maintenance’, section
’System Bootstrap’) for information how to force your unit to boot the secondary partition image.
7.4.2 Upgrading Viper 12A and 20A with version lower than 4.22.0
Upgrading a Viper 112A/212A or Viper 120A/220A from WeOS 4.21.2 (or earlier) to WeOS 4.22.0
(or later) will fail due to a limitation when handling larger WeOS images. Thus, when upgrading such
Viper products to WeOS 4.22.0 or later, first upgrade them to WeOS 4.21.3.