RX Drops on 3400cl

[Bruce Campbell_3]

What conditions result in "Drops RX" on the Procurve 3400cl product, running M.10.20 ? We have 6 3400cl switches, connected to a central 5400zl, all using gigabit over singlemode fibre. Intermapper shows the 6 3400cl switches getting excessive discards, 50-900 per minute, all 3400's do this at the same time. Then it stops. Traffic on the gig links is under 1% utilization typically. The RX Drops out of total are small, on the orderof 1 out of 1,500 Under what conditions does the 3400cl do an RX Drop ? Is it possible to log additional information regarding the reason for the drop ?

[Matt Hobbs]

Most likely these are false positives. I believe there will be a new firmware release coming out soon that will resolve most of these counter issues. Alternatively contact ProCurve support about this to see if you can get the updated firmware sooner.

[Bruce Campbell_3]

Thanks. I have a case with Procurve
support and will keep you posted.

I've determined
none of our 3400cl switches running
8.x code are showing RX Drops, and
all of the switches running 10.x
code are showing RX Drops. (we have
32 3400cl switches in total)

[Bruce Campbell_3]

Switch upgraded to M.10.28 shows 561 RX Drops after 30 minutes

Switch reverted to M.08.106 shows zero RX Drops after 30 minutes

[Matt Hobbs]

Bruce,

Do you have any ACL's on those ports? Can you possibly attach your config here and indicate which ports are incrementing the drops?

Matt

[Bruce Campbell_3]

no acls. On the switch below, ports 4,5 and
23 have indicated RX Drops. (it has only
been up 12 hours).


Running configuration:

; J4905A Configuration Editor; Created on release #M.10.28

hostname "xxxxxx"
snmp-server location "XX XXXX"
max-vlans 64
time timezone -300
no cdp run
no web-management
interface 1
name "xxxxxx"
exit
interface 2
name "xxxxxx"
exit
interface 3
name "xxxxxx"
exit
interface 4
name "xxxxxx"
exit
ip default-gateway w.x.y.z
sntp server a.b.c.d
timesync sntp
sntp unicast
logging facility local3
logging a.b.c.d
vlan 1
name "DEFAULT_VLAN"
ip address dhcp-bootp
no untagged 1-24
exit
vlan 18
name "xxxxxx"
no ip address
tagged 1-6,23-24
ip igmp
exit
vlan 24
name "xxxxx"
no ip address
tagged 1-6,23-24
ip igmp
exit
vlan 53
name "xxxxx"
tagged 1-6,23-24
exit
vlan 62
name "xxxxx"
no ip address
tagged 1-6,23-24
ip igmp
exit
vlan 3012
name "xxxxx"
ip address w.x.y.v 255.255.255.192
tagged 1-24
exit
ip authorized-managers a.b.c.0 255.255.255.0
ip authorized-managers a.b.d.0 255.255.255.0
ip authorized-managers a.b.e.0 255.255.255.0
ip authorized-managers a.b.f.0 255.255.255.0
spanning-tree
spanning-tree priority 1
password manager

[Bruce Campbell_3]

problem appears fixed in M.10.32