Adj resolve request: Failed to resolve

[michael.j.johnson1]

I have been receiving the following error for quiet some time now:

Jan 11 03:53:48.209: %ADJ-3-RESOLVE_REQ: Adj resolve request: Failed to resolve 10.91.123.247 Vlan123

I have attached the vlan and interface description.

NCJC_TelcoMDF_3750x_Stack#show vlan

VLAN Name Status Ports
---- -------------------------------- --------- -------------------------------
1 default active Gi1/0/11, Gi2/0/12
123 NCJC_Voice active Gi1/0/1, Gi1/0/2, Gi1/0/3, Gi1/0/4, Gi1/0/5, Gi1/0/6, Gi1/0/7, Gi1/0/8, Gi1/0/9, Gi1/0/10
Gi2/0/1, Gi2/0/2, Gi2/0/3, Gi2/0/4, Gi2/0/5, Gi2/0/6, Gi2/0/7, Gi2/0/8, Gi2/0/9, Gi2/0/10
Gi2/0/11
953 to_UZSW3_in_DC active
1002 fddi-default act/unsup
1003 token-ring-default act/unsup
1004 fddinet-default act/unsup
1005 trnet-default act/unsup

VLAN Type SAID MTU Parent RingNo BridgeNo Stp BrdgMode Trans1 Trans2
---- ----- ---------- ----- ------ ------ -------- ---- -------- ------ ------
1 enet 100001 1500 - - - - - 0 0
123 enet 100123 1500 - - - - - 0 0
953 enet 100953 1500 - - - - - 0 0
1002 fddi 101002 1500 - - - - - 0 0
1003 tr 101003 1500 - - - - - 0 0
1004 fdnet 101004 1500 - - - ieee - 0 0
1005 trnet 101005 1500 - - - ibm - 0 0

Remote SPAN VLANs
------------------------------------------------------------------------------

Primary Secondary Type Ports
------- --------- ----------------- ------------------------------------------

NCJC_TelcoMDF_3750x_Stack#sho int des
Interface Status Protocol Description
Vl1 admin down down
Vl123 up up "NCJC Voice"
Vl953 up up connection to uzsw3 in DC
Fa0 admin down down
Gi1/0/1 up up to Wake-NCJC-MDF-RTR
Gi1/0/2 up up to 91WAK-VG224-1
Gi1/0/3 up up to 91WAK-VG224-2
Gi1/0/4 up up to 91WAK-VG224-3
Gi1/0/5 up up Connect to NCJC-CM-Sub1-mega cluster
Gi1/0/6 down down
Gi1/0/7 down down
Gi1/0/8 down down
Gi1/0/9 down down
Gi1/0/10 down down
Gi1/0/11 down down
Gi1/0/12 up up Connect to uzsw3 in data center
Gi1/1/1 down down
Gi1/1/2 down down
Gi1/1/3 down down
Gi1/1/4 down down
Te1/1/1 down down
Te1/1/2 down down
Gi2/0/1 up up Connect to NCJC - sub 1 RSA mega cluster
Gi2/0/2 up up Connect to NCJC USC C240
Gi2/0/3 up up Connect to 91WAK-TelePres-VG3925.nccourts.org
Gi2/0/4 up up Connect to NCJC UCS_Vmware Server(4 logical connection)
Gi2/0/5 down down
Gi2/0/6 down down
Gi2/0/7 up up
Gi2/0/8 down down
Gi2/0/9 down down
Gi2/0/10 down down
Gi2/0/11 down down
Gi2/0/12 down down
Gi2/1/1 down down
Gi2/1/2 down down
Gi2/1/3 down down
Gi2/1/4 down down
Te2/1/1 down down
Te2/1/2 down down
Lo1 up up management interface

[Madhukrishnan Gopinathan Nair · Accepted solution]

Hello Michael,

Looks like you are hitting 

https://tools.cisco.com/bugsearch/bug/CSCuo89272/?reffering_site=dumpcr

%ADJ-3-RESOLVE_REQ displayed continuously after upgrade to 15.0(2)SE6

CSCuo89272

Description

Symptom:
after upgrade the IOS from 15.0(2)SE2 to 15.0(2)SE6, Below message displayed continuously.

----------
May 9 14:22:57.124 JST: %ADJ-3-RESOLVE_REQ: Adj resolve request: Failed to resolve 10.248.XX.XX Vlan301 May 9 14:23:11.125 JST: %ADJ-3-RESOLVE_REQ: Adj resolve request: Failed to resolve 10.248.XX.XX Vlan301 May 9 14:23:25.125 JST: %ADJ-3-RESOLVE_REQ: Adj resolve request: Failed to resolve 10.248.XX.XX Vlan301 May 9 14:23:39.125 JST: %ADJ-3-RESOLVE_REQ: Adj resolve request: Failed to resolve 10.248.XX.XX Vlan301
----------

We checked the show arp and found the entry for 10.248.XX.XX was incomplete status.

Conditions:
the ip address is incomplete status in show arp.

Workaround:
execute "no ip cef optimize neighbor resolution"

Further Problem Description:

Thanks,

Madhu.

[Madhukrishnan Gopinathan Nair · Accepted solution]

You should see these ip address in "sh ip arp" as incomplete.

Please check your running version of IOS is affected.

Sh version to check. If its affected, please try the work around as mentioned in the bug.

Workaround:
execute "no ip cef optimize neighbor resolution"

[elpollodiablo]

Fast forward 4.5 years and this is still a thing.

Version 15.2(7)E2 on a c3560x-8XPD-S getting same symptoms.  Workaround cleared that up.

[thbgodoy]

152-7.E3 still have the same problem.... workaround aplied. Thank you guys!

[DHOUHA BEN HAMZA]

Thks for Help

[Shaqxify]

Thank You So Much. God Bless You

[BrianSekleckiGE]

All:

There must be 100+ posts in the Cisco community about this error message, and the known-work-around.  

However, what Is missing from those messages:  Does the work-around impact performance?

Does anyone have any official communication from Cisco (or AB-Rockwell?)

In the mean time, I propose a second/alternative solution to simply filter out those messages from the log with a "discriminator" config, while leaving CEF optimization fully enabled.

!
!
no logging buffered
no logging console
!
no logging discriminator Filter_CEF_ErroneousWarning
logging discriminator Filter_CEF_ErroneousWarning facility drops ADJ severity drops 5 mnemonics drops RESOLVE_REQ_FAIL msg-body drops Adj
!
logging buffered discriminator Filter_CEF_ErroneousWarning 32786 informational
logging console discriminator Filter_CEF_ErroneousWarning debug
!
!

[bnidacoc]

A completely valid question Brian.  I've not seen a discussion around this pointless log message's "work-around" impact to the switch's performance.   I've read that CEF uses routines and cache tables in ASICs to build a switch's forwarding information base.  If a task isn't performed in an ASIC, it would seem that it would be performed in software/main general purpose CPU.  However, I thought that Cisco has told us over the decades that their product's performance is due to the use of their ASICs.

I recall the days when on 7500 series routers and Catalyst 6500s, performance issues were resolved with the Cisco TAC recommended global configuration command on 7500s, "ip cef".  Or TAC wanted to ensure CEF wasn't disabled on Cat 6500s.

[Bugleburner]

Be careful you fix the root cause, rather than just stopping the log messages.
For my issue, I saw the same messages alongside an incomplete arp entry. 

The issue was the firewall inside interface hadn't been plugged back in after switch RMA.

[BrianSekleckiGE]

Righto; normally when the IE-4000 (Catalyst 3750) is in Layer3 mode, and it cannot ARP on behalf of a client in a different VLAN, it causes this error message.  Normally on every other Cisco platform in history, this would be a silent failure, but on this platform it erroneously spews log messages.

Then the question of the work-around to fix the erroneous CEF log messages is to disable CEF globally, which is not good.   Fix the bloody code , Cisco >:}

[bnidacoc]

Exactly Brian.  Cisco needs to actually fix the code. 

This message used to never be there.  Cisco IOS software development introduced the overlogging condition which was never logged before, Cisco TAC then issues written solutions to just turn off a subset of CEF.  This is pathetic. 

When a host system (chatty windows servers) on a different network initiates communication with a host IP on that L3 switch's network and that host IP was a desktop taken out of service, window's chattiness will cause this condition endlessly because windows checks-in with windows clients ("are you still there?")

Anyone can recreate this issue by pinging a host IP they know is not actually present on that L3 switch. 

Window's chattiness and security/management scanning tools (IP sweeps) have been around since the 90s, but this Notification level log just showed up in the 2015-16 (or thereabouts).

This illustrates just how out of touch with the realities of a corporate network that the decision makers of Cisco's IOS Software management suite really are.

Just disable a CEF feature, TAC says; which Cisco has shipped CEF on by default since the early 2000s. Terrible.

[BrianSekleckiGE]

yes but filtering it in SYSLOG / Local Log with a discriminator fixes this until Cisco IOS is dead/replaced with IOS-XE:

!
!
no logging buffered
no logging console
!
no logging discriminator Filter_CEF_ErroneousWarning
logging discriminator Filter_CEF_ErroneousWarning facility drops ADJ severity drops 5 mnemonics drops RESOLVE_REQ_FAIL msg-body drops Adj
!
logging buffered discriminator Filter_CEF_ErroneousWarning 32786 informational
logging console discriminator Filter_CEF_ErroneousWarning debug
!

[dl_itnetwork]

Can anyone confirm if this is resolved in 15.2(7)E7? We have the same issue on a 3560-CX running 15.2(4)E6. 

Currently we are using the suggested discriminator as outlined above with success, but would prefer to not have to do this if possible.

[Darius_WWI]

The issue is still there in 15.2(7)E7. When the switch is L2-only you could use "no ip routing" to resolve this.