v1.5.2 Bug Reports and Comments

[mnitech]

Re multiple processes:

Just installed 1.5.2 on my WS-12-250-AC. Console says

ps axw | grep switch
1368 admin 3340 S switch -g
1371 admin 3340 S switch -g
1373 admin 3340 S {watchdog thread} switch -g
1376 admin 3340 S {i2c thread} switch -g
1377 admin 3340 S {poe thread} switch -g
1607 admin 3140 S switch -s
1679 admin 3140 S switch -s
2192 admin 2680 S grep switch

To which process do you refer?

[Stephen]

If you suspect this is happening it is really better to check it with top. Top moves the active processes to the top of the list and for the issue I was referring too to present itself it would show 2 'switch' processes at the top.

The switch process itself fork's a few time's so there are normally suppose to be multiple instance's running. This detail isn't necessary to know if you're just using top though. What is happening that lead's you to suspect you have this issue?

[ahteran]

Hello everyone
WS-8-150-DC NTP not sync
I try everyhing
Static ip
Dynamic ip
DNS
Different ntp server web addresses and ip numbers
But the time not set warning has not been removed
Cannot be assigned a time period to ports

[sakita]

Can the switch ping the NTP server? On the tools tab there is a ping tool which can be used to verify that the switch can get to the NTP server...

[sirhc]

If the NTP server you want to use is on the public net but your switch does not have access to public net as it might have an invalid IP then you can not.

Most WISPs run an internal NTP server so they can get to it via their invalid IPs.

Not sure why you put this in Bug for v1.5.2 as this is NOT a bug but a question on what your missing as to why you can not.

Think about it, 70,000+ switches all over the world and this is a common thing so what are the odds this is a BUG and your the only one to find it, people would be SCREAMING ?

I run these switches at my WISP and I know NTP works on all versions before I post them.

If your switch is accessible from WWW such as having a valid IP then make sure you use the Access list to prevent hacks. If you have port mappings then implement an access control list on your router to prevent just anyone from accessing it.

[coreinput]

Upgraded from 1.5.1 running since Feb 2019 and everything went smooth (and no POE loss).
Hats off to the team including beta testers for making this a good release.
Thanks!

p.s. NTP worked for me on the upgrade:

Code: Select all

Dec 31 19:00:57 system: starting ntpclient
May 1 09:36:57 system: time set by NTP server


[Stephen]

Ah I see,

This was actually a feature request, to allow people to select which protocol was being searched for instead of all of them being searched for regardless of which protocols where enabled.

I can add an option on the next release to allow you to disable whether or not the switch is broadcasting on the enabled discovery protocol if you want.

1.5.3rc1 has this added in.

[aaron@futuretk.com]

WS-26-500-DC - Packet Loss after upgrading to 1.5.2.

I've got two brand new WS-26-500-DC switches on the bench. Both came out of the box with 1.5.0. After upgrading to 1.5.2 both are showing packet loss when I ping them locally. Around 10% loss, sometimes a little more. Downgrading them back to 1.5.0 fixes the problem. Upgrading again to 1.5.2 brings the issue back. No other configuration has been done on either switch.

[aaron@futuretk.com]

After some further testing it appears that this is related to the switch boot up process. After upgrading to 1.5.2 I show packet loss from the time the first ping comes back, until about 3 minutes after first ping. After that it seems to clear up. This does not happen in 1.5.0 under the same conditions.

[mskteamwl]

Hi,

We've made a benchtest with JDSU 1Gbps devices and the Switch RTD increase drastically until we experience packet lost. Near 1800 us.

This take about 2 minutes of stress test.

Any idea about the problem?

--- SNIP ---

Ok found a solution track in this post.

https://forum.netonix.com/viewtopic.php?f=17&t=1775Keep you in touch