RNRP Interface blocked for over 60min
We had a Network issue for 6s, but the Interface was blocked for 62min.
RNRP Log:
31/10*16:49:23.258 ConfError: Interface unblocked on area=1 path=0,
31/10*16:49:24.026 ConfError: Interface blocked after detected network loop on area=1 path=0
31/10*17:51:07.530 ConfError: Interface unblocked on area=1 path=0,
One Point is the Network issue. (lost one port of Portchannel). Why is the Interface blocked? And why is the Interface blocked for over 60min?
(Problem was only on Server Client LAN)
Our System:
6.0.3.1
System full virtual over 4 Esxi.
thx
RNRP Log:
31/10*16:49:23.258 ConfError: Interface unblocked on area=1 path=0,
31/10*16:49:24.026 ConfError: Interface blocked after detected network loop on area=1 path=0
31/10*17:51:07.530 ConfError: Interface unblocked on area=1 path=0,
One Point is the Network issue. (lost one port of Portchannel). Why is the Interface blocked? And why is the Interface blocked for over 60min?
(Problem was only on Server Client LAN)
Our System:
6.0.3.1
System full virtual over 4 Esxi.
thx
Asked 2 months ago
Answers
2
The loop must have endured for more than 7 seconds.
Upon sensing same routing telegram from any peer too many times (actually 13 times) RNRP will perform a short interface block of 7 seconds. If the loop continues after 7 seconds, a 60 minute block is made if there is a redundant path available.
The message you pasted is the 2nd message, the long block.
This 1 hour block can be released in Windows nodes by using the RNRP Utility, even remotely, just remember to issue the release command via an unaffected and open path. You can also specifiy a multicast address to unlock several computers on the same subnet/area as the unlocking computer is connected to. Use 239.239.239.area x 4 + path, eg 239.239.239.81 to issue command to all nodes on area 20, via secondary path.
Loops on process control networks is a serious matter. Hard loops risk to overload controllers. Many ABB PLCs will perform automatic shutdown/halt if exposed to overload. Find out what causes the loop and resolve it. Spanning Tree (including RSTPJ is too slow to react as a first line of defense. ABB recommend using a proprietary L2 redundancy protocol in rings, eg MRP, TurboRing, FRNT, HIPER Ring, etc. Rings are cheaper but offer less protection than a truly redundant network using two physically isolated network paths. STP can then be used on non-ring ports as a second line of defense against human errors. Likely, RNRP will engage before STP has reacted, but communication should be restored after Forward Delay (up to 30 seconds in STP defaults). Remember, a short hard burst may be enough to halt a controller before RNRP or the AC 800M Storm Filter react; don’t trust STP to save you from a shutdown.
Search ABB Library for a Technical Description detailing RNRP Loop Detection & Protection and the AC 800M Storm Filter.
Upon sensing same routing telegram from any peer too many times (actually 13 times) RNRP will perform a short interface block of 7 seconds. If the loop continues after 7 seconds, a 60 minute block is made if there is a redundant path available.
The message you pasted is the 2nd message, the long block.
This 1 hour block can be released in Windows nodes by using the RNRP Utility, even remotely, just remember to issue the release command via an unaffected and open path. You can also specifiy a multicast address to unlock several computers on the same subnet/area as the unlocking computer is connected to. Use 239.239.239.area x 4 + path, eg 239.239.239.81 to issue command to all nodes on area 20, via secondary path.
Loops on process control networks is a serious matter. Hard loops risk to overload controllers. Many ABB PLCs will perform automatic shutdown/halt if exposed to overload. Find out what causes the loop and resolve it. Spanning Tree (including RSTPJ is too slow to react as a first line of defense. ABB recommend using a proprietary L2 redundancy protocol in rings, eg MRP, TurboRing, FRNT, HIPER Ring, etc. Rings are cheaper but offer less protection than a truly redundant network using two physically isolated network paths. STP can then be used on non-ring ports as a second line of defense against human errors. Likely, RNRP will engage before STP has reacted, but communication should be restored after Forward Delay (up to 30 seconds in STP defaults). Remember, a short hard burst may be enough to halt a controller before RNRP or the AC 800M Storm Filter react; don’t trust STP to save you from a shutdown.
Search ABB Library for a Technical Description detailing RNRP Loop Detection & Protection and the AC 800M Storm Filter.
by ahuerlimann Rank: 319 on 11/30/2018 1:46:59 AM | Like (0) | Report
Sorry for double post. Got a service unavailable message from forum...
Add new comment