v1.4.5rcX Bug Reports and Comments

[sirhc]

Just configuring a Metrolink-Link using *one* Netonix Switch. There seem to be a loop protection mechanism which makes both unavailable. This even happens doing both Metrolinqs in different Vlans.

There is no bug with Loop Protection

Here is a link to a search for (click on magnifying glass in search box):
Search for keywords: +Loop +Protection
Author: sirhc

search.php?keywords=%2Bloop+%2Bprotection&terms=all&author=sirhc&sc=1&sf=all&sr=posts&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

Read down thorough the posts in the link above to learn more

You can disable Loop Protection on the Device/Configuration Tab but something is leaking BPDU packets, I would find out what.

A BPDU packet is leaving 1 port and finding its way back to the same or another port - IT SHOULD NOT HAPPEN UNLESS THERE IS A LEAK OR A LOOP SOMEWHERE

viewtopic.php?f=17&t=1459&p=12102&hilit=Loop+Protection#p12102

Loop Protection is not like RSTP.

It sends a special packet out each port every second and if the packet comes back even to the same port or another port it shuts the sending or receiving port down for 180 seconds.

So if a customer or another piece of equipment further out on your network has a leaky VLAN or a loop it will cause this to happen.

Leaky VLANs or equipment can cause false loop protections.

[mike99]

Just configuring a Metrolink-Link using *one* Netonix Switch. There seem to be a loop protection mechanism which makes both unavailable. This even happens doing both Metrolinqs in different Vlans.

We had a metrolinq link with a Netonix (WS-12-250-AC with 1.4.2) powering them at both side and STP was working fine, port always forwarding, but only 60 Ghz was connected and 5 Ghz was disable. Do you have both frequency connect at the same time ? Maybe try to disable 5 Ghz. Also, have you upgrade the metrolinq to last firmware ? We had several bug until firmware was upgraded.

[mayheart]

First time running a DC switch, is it normal for it to be around 45-50% cpu all the time? I'm running 1.4.5rc4 on it. BTW this is running off a -48V telecom plant with a Meanwell 350V DCtoDC converter. Working well so far.

1342 root 20 0 3272 1120 572 S 46.6 0.9 322:35.43 switch
1354 root RT 0 3272 1120 572 S 5.2 0.9 53:49.97 i2c thread

There is maybe less than a mbit worth of traffic with tagged management going through it.

[sirhc]

I would suggest running v1.4.5rc6 but this has been answered many times before.

There is maybe less than a mbit worth of traffic with tagged management going through it.

The CPU is there to run the UI/CLI/Daemons and configures the "Switch Core", the Switch Core is a separate component and handles all packet switching independently of the CPU running Linux.

This is NOT a software switch but rather a Switch Core the same way all normal switches function.

The WISP Switch uses the VSC-742X Vitesse Enterprise Switch Core.
The TOUGH Switch uses the BCM-5311X Broadcom SOHO Switch Core.
viewtopic.php?f=6&t=185&p=474&hilit=broadcom#p474

I did a Search our forum and found this answer:
viewtopic.php?f=17&t=2106&p=15348&hilit=+cpu+capacity#p15348

The higher CPU is from the added graphs.

You do notice that the Device/Status Tab now has 24 hour graphs that you guys asked for.

The more features you request the more CPU will get used.

The CPU has no bearing on switch capacity, all the CPU does is run the UI/CLI/SNMP/Stats/Daemons.

It spikes when you first log in as it is busy setting up the information the UI need such as Logs and stats then drops down if you watch it. It can also fluctuate higher for periods of time when it is on a larger flat segment if you have discover Tab or protocols enabled as it is finding devices.

[sirhc]

FIXED[b]/CHANGED[/b]
- Fixed voltage fluctuation that affected a small percentage of WS-6-MINI, WS-8-150-AC, WS-8-150-DC - rc1
- Updated SSH daemon to get more modern key exchange algorithm - rc1
- Watchdog no longer checks for link on port to be bounced - rc1
- Ignore the SFP ports when checking Priority values - rc4
- Changed throttle limits between switch core and CPU - rc4
- Fixed web UI login only using the first 8 characters of the password - rc4
- Removed Address from MODBUS configuration - rc4
- Fixed reset button on WS-6-MINI broken since v1.4.1 - rc6

ENHANCEMENTS
- Added option to bounce link instead of POE for watchdog rules - rc1
- Added support for 1000BASE-BX SFP modules (including Fiberstore SFP-GE-BX) - rc2
- Added ability to configure QoS based on packet headers (MAC, IP, TCP ports, etc) - rc6
- Added support for Mikrotik S-RJ01 SFP module - rc6
- Show the SFP model number in UI - rc6
- Skip PoE watchdog if PoE is not turned on for a port - rc7
- Changed default Revert Timer value from 60 to 180 seconds - rc7
- Added range for DSCP and PCP values in QoS rules - rc7
- Removed DSCP and PCP mappings from QoS config - rc7

KNOWN ISSUES
- Some language templates need help - please contact us to help
- UI broken on mobile devices - will be fixed in - rc8

Released 10/21/2016 - rc7
Released 10/18/2016 - rc6
Released 10/10/2016 - rc4
Released 9/21/2016 - rc2
Released 9/19/2016 - rc1

[ste]

Just configuring a Metrolink-Link using *one* Netonix Switch. There seem to be a loop protection mechanism which makes both unavailable. This even happens doing both Metrolinqs in different Vlans.

We had a metrolinq link with a Netonix (WS-12-250-AC with 1.4.2) powering them at both side and STP was working fine, port always forwarding, but only 60 Ghz was connected and 5 Ghz was disable. Do you have both frequency connect at the same time ? Maybe try to disable 5 Ghz. Also, have you upgrade the metrolinq to last firmware ? We had several bug until firmware was upgraded.

Firmware of metrolinq is the newest and 5 GHZ was disabled. As soon as 60ghz connects I can't ping the metrolinqs.
Loop protection on netonix is disabled. I created a second vlan on the netonix switch and configured the port of one of the metrolinqs as access port into this vlan. As soon as I unplug this metrolinq and power it with a poe adapter both are reachable.

[lligetfa]

If it works with the PoE adapter but not powered by the switch, then it is one of two things. 1. An overcrimped cable is making a better connection into the PoE brick than it makes into the switch, or 2. There is something funky effect on the switch due to the power being drawn from the switch. Again, an overcrimp could mean the pair doubling on the power pairs isn't happening and all the current might be carried by fewer wires.

[mike99]

Firmware of metrolinq is the newest and 5 GHZ was disabled. As soon as 60ghz connects I can't ping the metrolinqs.
Loop protection on netonix is disabled. I created a second vlan on the netonix switch and configured the port of one of the metrolinqs as access port into this vlan. As soon as I unplug this metrolinq and power it with a poe adapter both are reachable.

I would try the link between 2 switchs in lab just to make sure the link is the problem. If so, maybe write to Ignitenet, their support team seem reactive from my experience even if I didn't have to ask technical questions.

Like I already writen, we had one for around a month or two in production between 2 switchs and it was working fine but we replaced it because it was dropping at each rain and use it instead for a customer P2P.

[mike99]

You already can optionally add a subnet to the src IP.

What the syntax for subnet ? I was not able to configure it.
I tryed:

192.168.1.0/24
192.168.1.0-192.168.1.255
192.168.1.1-192.168.1.254
192.168.1.0 255.255.255.0

DSCP range is now support. I hope to do more QoS test this week. I also saw that default queue by port was remove.

[Eric Stern]

192.168.1.0/255.255.255.0

If you mouse over the field it will show you that.

Default queue by port was removed because you can do that with a rule now.