× NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Orbi WiFi 7 RBE973

Re: S3300-28X: behaviour after managment unit conflict.

remosito
Aspirant

S3300-28X: behaviour after managment unit conflict.

If a managment unit conflict is detected (for whatever reason). And a new election is held and a new management unit is assigned. According to log files other stack member will reset and rejoin. 

 

Now my question is "what does reset mean here?". An actual reboot?

 

 

(additional question about stacking. if I have 3 switches in line 1-2-3 (no direct connection between 1 and 3). with 1 being master.

 

if link 1-2 goes down. master will disappear to switches 2 and 3. Am I right in assuming 2 and 3 will determine a new master between the two of them. And once link 1-2 comes back. there will be conflict between old master 1 and new master from 2/3? 

 

what happens if middle switch 2 goes down? will 3 elect itself as master of it's now one switch stack. and once 2 comes back we have conflict between 1 and 3?

)

Model: S3300-28X (GS728TX)|ProSafe 24 ports stackable smart switches
Message 1 of 10

Accepted Solutions
DaneA
NETGEAR Employee Retired

Re: S3300-28X: behaviour after managment unit conflict.

@remosito,

 

I got a feedback from the higher tier of NETGEAR Support that the word "reset" in the logs you have posted means reboot. From the logs, it means that it is a reboot of the stack member which is unit 2. 

 

 

Regards,

 

DaneA

NETGEAR Community Team

View solution in original post

Message 9 of 10

All Replies
remosito
Aspirant

Re: S3300-28X: behaviour after managment unit conflict.

We would really appreciate a clarification as our stack has been acting up with reboots of switches.

 

If my logs say 

 

Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(6588) 699 %% Reset initiated on unit 2, reason: Stack manager conflict resolution
Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(5347) 698 %% Management unit conflict was detected and this unit has lost the manager election. All the units in this stack will reset and join the winner stack manager 8:bd:43:6e:75:d0.

Is a reset an actual reboot of the whole switch? Or just a reset of the stack that is done without a reboot?

Model: S3300-28X (GS728TX)|ProSafe 24 ports stackable smart switches
Message 2 of 10
DaneA
NETGEAR Employee Retired

Re: S3300-28X: behaviour after managment unit conflict.

@remosito,

 

Welcome to the community! 🙂 

 

We would really appreciate a clarification as our stack has been acting up with reboots of switches.

 

If my logs say 

 

Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(6588) 699 %% Reset initiated on unit 2, reason: Stack manager conflict resolution Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(5347) 698 %% Management unit conflict was detected and this unit has lost the manager election. All the units in this stack will reset and join the winner stack manager 8:bd:43:6e:75:d0.

 

Is a reset an actual reboot of the whole switch? Or just a reset of the stack that is done without a reboot?

Let me inquire this to the higher tier of NETGEAR Support.  I will post here once I got a feedback. 

 

(additional question about stacking. if I have 3 switches in line 1-2-3 (no direct connection between 1 and 3). with 1 being master.

 

if link 1-2 goes down. master will disappear to switches 2 and 3. Am I right in assuming 2 and 3 will determine a new master between the two of them. And once link 1-2 comes back. there will be conflict between old master 1 and new master from 2/3? 

 

what happens if middle switch 2 goes down? will 3 elect itself as master of it's now one switch stack. and once 2 comes back we have conflict between 1 and 3?

Here are the factors determine which switch is elected as the stack master:

 

a. The switch that is manager always has priority to retain the role of master
b. Assigned priority
c. MAC address

 

Be reminded that the priority of the switch is how the stack decides what switch will be the master (stack manager).  The highest priority we can give a switch is 15, though keep in mind that you can have one switch with a priority of 1 and the other switches not have any priority configured, then the switch with priority of 1 will be the master. But keep in mind that the switch with the next highest priority will be the backup if the master fails.  If the priority for the switches in the stack has not been configured or they have the same priority, then the switch with the highest MAC address value will be elected as the master.  If you give a priority of 0 to a switch it will never be considered for management.

 

 

Regards,

 

DaneA

NETGEAR Community Team

Message 3 of 10
DaneA
NETGEAR Employee Retired

Re: S3300-28X: behaviour after managment unit conflict.

@remosito,

 

I just got a response from the higher tier of NETGEAR Support.  As per the higher tier of NETGEAR Support, it might be possible that it is a reboot of the individual stack member.  For us to confirm, kindly check the uptime of that stack member.  Go to System > Stacking > Basic > Stack Configuration.  

 

 

Regards,

 

DaneA

NETGEAR Community Team

Message 4 of 10
remosito
Aspirant

Re: S3300-28X: behaviour after managment unit conflict.

thanks for reaching out to higher tier netgear support.

 

My Problem is that mine are rebooting sometimes when I do work on them and touch cables. (I know from uptime btw that they are rebooting).

 

The question is. Are they rebooting because something in our inwall cabling is only slightly screwed and stack members loose connection between themselves for short duration when I touch cabling. And it is default behaviour of my switches to reboot when the shortly separated members rejoin (I see the RESET entry in the log). Which would make it rather obvious where exactly my cabling problem is (namely the one connecting stack members on different floors).

 

Or is  one or multiple of our inwall cabling totally screwed and my switches are rebooting because touching cables results in "illegal/unhealthy" currents to reach the switch resulting in this one making an emergency reboot? In which case cabling problem could be on any of the inwall ethernet connections to workplaces.

 

We already know it's the inwall cabling because the first thing we did is buy another S3300 switch and replace it. Didn't help. The second thing we did is replace all short cables from switch to plugs leading to inwall cabling. Didn't help either.

 

I just need to know if the logs say reset.

 

Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(6588) 699 %% Reset initiated on unit 2, reason: Stack manager conflict resolution
Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(5347) 698 %% Management unit conflict was detected and this unit has lost the manager election. All the units in this stack will reset and join the winner stack manager 8:bd:43:6e:75:d0.

what actually is meant is reboot. If you could kindly reask higher tier netgear support. That would be great. "Might" is not helpful really. Somebody ought to know if log entry of reset is actual reboot.

Message 5 of 10
DaneA
NETGEAR Employee Retired

Re: S3300-28X: behaviour after managment unit conflict.

@remosito,

 

My Problem is that mine are rebooting sometimes when I do work on them and touch cables. (I know from uptime btw that they are rebooting).

 

The question is. Are they rebooting because something in our inwall cabling is only slightly screwed and stack members loose connection between themselves for short duration when I touch cabling. And it is default behaviour of my switches to reboot when the shortly separated members rejoin (I see the RESET entry in the log). Which would make it rather obvious where exactly my cabling problem is (namely the one connecting stack members on different floors).

 

Or is  one or multiple of our inwall cabling totally screwed and my switches are rebooting because touching cables results in "illegal/unhealthy" currents to reach the switch resulting in this one making an emergency reboot? In which case cabling problem could be on any of the inwall ethernet connections to workplaces.

 

We already know it's the inwall cabling because the first thing we did is buy another S3300 switch and replace it. Didn't help. The second thing we did is replace all short cables from switch to plugs leading to inwall cabling. Didn't help either.

With regard to this, it would be best that you open a chat or online support ticket with NETGEAR Support at anytime.  Kindly state your concern and you may indicate a link to this forum thread as reference.  In this way, a thorough investigation will conducted on your switch stack.  

 

 

Regards,

 

DaneA

NETGEAR Community Team

Message 6 of 10
remosito
Aspirant

Re: S3300-28X: behaviour after managment unit conflict.

I will contact support that way. Thank you for your help.

 

But tbh I don't really need a thorough investigation of my stack. I just need to know when a S3300 writes these log entries and talks about RESET. If what he actually does is a REBOOT.

 

then everything would be already clear.

 

Will come back after support answered the question and post it here.

Message 7 of 10
remosito
Aspirant

Re: S3300-28X: behaviour after managment unit conflict.

Hi again,

 

tried contacting support, but after entering Serial Number and Date of sale it tells me my 90 day support ended already. And I have to buy some support contract.

 

Now I might consider doing that if I actually had a real support case! I don't.

 

All I wanna know is 

 

if I see this in my logs:

Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(6588) 699 %% Reset initiated on unit 2, reason: Stack manager conflict resolution
Keller-2 UNITMGR[unitMgrTask]: unitmgr.c(5347) 698 %% Management unit conflict was detected and this unit has lost the manager election. All the units in this stack will reset and join the winner stack manager 8:bd:43:6e:75:d0

does that Reset mean actual reboot of units.

 

I really shouldn't have to buy a support contract for what is in my opinion essentially information that should be publicly available in either a manual or a FAQ. 

 

Could you please reask the support guys on your end to confirm reset here means  "it is a reboot" or "it is not a reboot".

Message 8 of 10
DaneA
NETGEAR Employee Retired

Re: S3300-28X: behaviour after managment unit conflict.

@remosito,

 

I got a feedback from the higher tier of NETGEAR Support that the word "reset" in the logs you have posted means reboot. From the logs, it means that it is a reboot of the stack member which is unit 2. 

 

 

Regards,

 

DaneA

NETGEAR Community Team

Message 9 of 10
remosito
Aspirant

Re: S3300-28X: behaviour after managment unit conflict.

Hi DaneA,

 

totally forgot to reply to you and thank you for your effort to get to the bottom of this! Very much appreciated!

 

cheers 

 

remo

Message 10 of 10
Top Contributors
Discussion stats
  • 9 replies
  • 2302 views
  • 1 kudo
  • 2 in conversation
Announcements