Need help with OPERATOR.LOG Message.

[cyberunlimited]

One of our DS-25 Alpha’s that we use for our SCADA System started to show issues with the Mailbox Router and one of the tasks actually had a RWMBX Message when I did a Show System Command. Also there were numerous logs of this in the OPERATOR.LOG File:

%%%%%%%%%%% OPCOM 27-OCT-2020 09:14:50.69 %%%%%%%%%%%
Message from user SYSTEM on CVHQB
Event: Unavailable User Buffer from: Node LOCAL:.CVHQB CSMA-CD Station CSMACD-0,
at: 2020-10-27-09:14:50.694-08:00Iinf
eventUid DE01860B-1834-11EB-BF0C-AA0004000C04
entityUid 4D8FADB1-2595-11EA-8698-AA0004000C04
streamUid 4FFE5BAF-2595-11EA-87E7-AA0004000C04

Have you guys ever seen this issue before?

Thanks.

[Volker Halle]

From the DECnet-Plus NCL Langeuage Reference Manual:

unavailable user buffers

Number of times no user buffer was available for an incoming frame that passed all filtering for the port. Used with the pdus received count, this counter can indicate the rate of user buffer receive problems.

There seems to be a resource problem on this DS25 or too much incoming packets on the CSMACD-0 network interface.

A process enters RWMBX state, if it tries to WRITE to a mailbox, which is already full. The process emptying/reading the mailbox is either too slow or stuck.

Volker.

[cyberunlimited]

Thanks for the help. 

[Volker Halle]

You can look at the 'Unavailabe User Buffer' counters with SDA, to find out the network protocol,  those events are related to:

$ ANALYZE/SYSTEM

SDA> SHOW LAN

...

-- EWAx xx-xx (...) Counters Information

...

Unavail user buffer 0

...

Volker.

[cyberunlimited]

I see information about various Devices but I don't see anything that references the unavailable user buffer. I never see the

Unavail user buffer 0

Thanks. 

[Volker Halle]

$ ANALYZE/SYSTEM

SDA> SHOW LAN/COUNTER

will show 'Unavail user buffer' for the LAN interface itself and all the LAN devices for the different protocols.. Works for me on OpenVMS Alpha V8.2 and even on OpenVMS VAX V7.3

Volker.

[cyberunlimited]

Yes, that worked. 

For the EIA Counters, here's what showed up. The Unavailable User Buffers shows 559993. 

LAN Data Structures
-------------------
-- EIA Counters Information 30-OCT-2020 09:24:38 --

Octets received 4745310707964 Octets sent 647192707857
PDUs received 4672985531 PDUs sent 1865371435
Mcast octets received 4254256590836 Mcast octets sent 371135259159
Mcast PDUs received 2955341236 Mcast PDUs sent 261228424
Unrec indiv dest PDUs 0 PDUs sent, deferred 0
Unrec mcast dest PDUs 693388384 PDUs sent, one coll 0
Data overruns 0 PDUs sent, mul coll 0
Unavail station buffs 0 Excessive collisions 0
Unavail user buffers 559993 Late collisions 0
CRC errors 0 Carrier check failure 0
Alignment errors 0 Last carrier failure None
Rcv data length err 1 Coll detect chk fail 0
Frame size errors 0 Short circuit failure 0
Frames too long 0 Open circuit failure 0
Seconds since zeroed 26913369 Transmits too long 0
Station failures 0 Send data length err 0
No work transmits 0 Ring avail transitions 2
Buffer_Addr transmits 0 Ring unavail transitions 1
SVAPTE/BOFF transmits 0 Loopback sent 0
Global page transmits 0 System ID sent 90262
Bad PTE transmits 0 ReqCounters sent 0
Restart pending counter 0 Internal counters size 168
+00 CFG cmd buffer VA 84F3C000 +50 Device resets 1
+04 Cmd buffer list VA 84F3C020 +54 CXB transmits 300094302
+08 Receive ring VA 84F40020 +58 Inits nomap 4096
+0C Stat buffer VA 84F43020 +5C Inits mapped 1
+10 LSB size 8120 +60 Map regs allocated 0
+14 Driver version 43 +64 Rcv buffers mapped 198553661
+18 Driver flags 00003268 +68 Xmt segments mapped 496276662
+1C Device ID 00CE0080 +6C Page boundary xmts 24855384
+20 LAN_FLAGS 00000000 +70 PTE to PFN translates 0
+24 Soft errors 0 +74 PTE to PFN 2nd page 0
+28 Hardware errors 0 +78 Chained VCRP xmits 93208919
+2C Duplex mismatches 0 +7C Too few segment xmits 0
+30 Link state 3 +80 Transmit copies 7416
+34 Link transitions 3 +84 Zero length DCBEs 0
+38 Nway resets 0 +88 Transmit timeouts 0
+3C Nway state 00000000 +8C Transmits queued 0
+40 Reforks 10788 +90 Control commands done 1377109
+44 Cmd/Rcv unit state 00000050 +94 Receive lockups 0
+48 Device interrupts 2102843711 +98 Duplex mismatch time 0
+4C Last SCB read 1023410256 +9C Generic (or unused) 00000000

[Volker Halle]

Keep entering <RETURN> and you will see 'EIAx (protocol) Counters Information...' for each of the various network protocols active on that LAN interface. And each EIAx device also has 'Unavail user buffer' counters. Look for EIAx devices with a high count of unavailable user buffers...

SDA> SHOW LAN will also show the counter information, but you need to press <RETURN> much more often.

Volker.