Original Message:
Sent: Sep 13, 2023 06:01 AM
From: frmeunier
Subject: Device attached to IRF member losing network
You say "I have a server attached to the satellite switch (10.10.14.13) and the server never looses a single ping"
but the ping to the switch is unstable, right ? the machines behind the sat switch are OK, but not the switch itself ?
well, it may mean that it's probably not a stack issue, except if the hashing algorithm (that will distribute the frames through the splitted 4x10G irf link) has a problem working with that (trying to send the frame to lane 2, but frame not correctly received ?)
Did you try to connect another switch as satellite (just to have another mac-address and probably hit another hashing result) ?
Also :
display interface Ten-GigabitEthernet3/0/49:1
display interface Ten-GigabitEthernet3/0/49:2
display interface Ten-GigabitEthernet3/0/49:3
display interface Ten-GigabitEthernet3/0/49:4
display interface Ten-GigabitEthernet3/0/51:1
display interface Ten-GigabitEthernet3/0/51:2
display interface Ten-GigabitEthernet3/0/51:3
display interface Ten-GigabitEthernet3/0/51:4
and check for strange values (counters, errors, state)
------------------------------
Frederic MEUNIER
Original Message:
Sent: Sep 11, 2023 09:47 AM
From: spgsitsupport
Subject: Device attached to IRF member losing network
Isn't the split (40 to 40 via 4x10) supposed to be exactly the same, 40G ?
I could not find any info about it anywhere (what supported, what is not)
I have a server attached to the satellite switch (10.10.14.13) and the server never looses a single ping
It is only the 2920 stack that seems to be affected
Seb
------------------------------
spgsitsupport
Original Message:
Sent: Sep 11, 2023 08:53 AM
From: FREDERIC MEUNIER
Subject: Device attached to IRF member losing network
wait ! some stack links are 40G and some are 4x10G (splitted from 40G) ?
From my point of view, it's not a supported configuration : all links should be the same sort.
------------------------------
Frederic MEUNIER
Original Message:
Sent: Sep 11, 2023 08:15 AM
From: spgsitsupport
Subject: Device attached to IRF member losing network
Ofcourse MAC address does not change, as it is static on remote switch
If I attach the switch to member 1 or 2 then I never ever have an issue (the issue only exists when it is on the member 3 which is part of the IRF via OM4 fibre)
Only ports 49 & 51 are used in every unit for IRF purposes
[HPE5900-SR1]dis irfMemberID Role Priority CPU-Mac Description *+1 Master 32 f010-90db-7402 --- 2 Standby 30 f010-90db-7403 --- 3 Standby 20 f010-90db-7404 --- 4 Standby 10 f010-90db-7405 ----------------------------------------------------- * indicates the device is the master. + indicates the device through which the user logs in. The bridge MAC of the IRF is: e8f7-2451-3464 Auto upgrade : yes Mac persistent : 6 min Domain ID : 0 IRF mode : normal
[HPE5900-SR1]dis irf topology Topology Info ------------------------------------------------------------------------- IRF-Port1 IRF-Port2 MemberID Link neighbor Link neighbor Belong To 1 UP 4 UP 2 f010-90db-7402 4 UP 3 UP 1 f010-90db-7402 3 UP 2 UP 4 f010-90db-7402 2 UP 1 UP 3 f010-90db-7402
[HPE5900-SR1]dis irf linkMember 1 IRF Port Interface Status 1 FortyGigE1/0/49 UP FortyGigE1/0/50 DOWN 2 FortyGigE1/0/51 UP FortyGigE1/0/52 DOWNMember 2 IRF Port Interface Status 1 FortyGigE2/0/49 UP FortyGigE2/0/50 DOWN 2 FortyGigE2/0/52 DOWN Ten-GigabitEthernet2/0/51:1 UP Ten-GigabitEthernet2/0/51:2 UP Ten-GigabitEthernet2/0/51:3 UP Ten-GigabitEthernet2/0/51:4 UPMember 3 IRF Port Interface Status 1 Ten-GigabitEthernet3/0/49:1 UP Ten-GigabitEthernet3/0/49:2 UP Ten-GigabitEthernet3/0/49:3 UP Ten-GigabitEthernet3/0/49:4 UP 2 Ten-GigabitEthernet3/0/51:1 UP Ten-GigabitEthernet3/0/51:2 UP Ten-GigabitEthernet3/0/51:3 UP Ten-GigabitEthernet3/0/51:4 UPMember 4 IRF Port Interface Status 1 FortyGigE4/0/50 DOWN Ten-GigabitEthernet4/0/49:1 UP Ten-GigabitEthernet4/0/49:2 UP Ten-GigabitEthernet4/0/49:3 UP Ten-GigabitEthernet4/0/49:4 UP 2 FortyGigE4/0/51 UP FortyGigE4/0/52 DOWN[HPE5900-SR1]dis irf linkMember 1 IRF Port Interface Status 1 FortyGigE1/0/49 UP FortyGigE1/0/50 DOWN 2 FortyGigE1/0/51 UP FortyGigE1/0/52 DOWNMember 2 IRF Port Interface Status 1 FortyGigE2/0/49 UP FortyGigE2/0/50 DOWN 2 FortyGigE2/0/52 DOWN Ten-GigabitEthernet2/0/51:1 UP Ten-GigabitEthernet2/0/51:2 UP Ten-GigabitEthernet2/0/51:3 UP Ten-GigabitEthernet2/0/51:4 UPMember 3 IRF Port Interface Status 1 Ten-GigabitEthernet3/0/49:1 UP Ten-GigabitEthernet3/0/49:2 UP Ten-GigabitEthernet3/0/49:3 UP Ten-GigabitEthernet3/0/49:4 UP 2 Ten-GigabitEthernet3/0/51:1 UP Ten-GigabitEthernet3/0/51:2 UP Ten-GigabitEthernet3/0/51:3 UP Ten-GigabitEthernet3/0/51:4 UPMember 4 IRF Port Interface Status 1 FortyGigE4/0/50 DOWN Ten-GigabitEthernet4/0/49:1 UP Ten-GigabitEthernet4/0/49:2 UP Ten-GigabitEthernet4/0/49:3 UP Ten-GigabitEthernet4/0/49:4 UP 2 FortyGigE4/0/51 UP FortyGigE4/0/52 DOWN
No IRF error in log
The switch attached to XGE3/0/10 simply disappears every few moments
[HPE5900-SR1]display mac-address 3863-bb52-348bMAC Address VLAN ID State Port/NickName Aging3863-bb52-348b 14 Learned XGE3/0/10 Y[HPE5900-SR1]display mac-address 3863-bb52-348bMAC Address VLAN ID State Port/NickName Aging[HPE5900-SR1]
------------------------------
spgsitsupport
Original Message:
Sent: Sep 11, 2023 07:35 AM
From: FREDERIC MEUNIER
Subject: Device attached to IRF member losing network
Hello
I shall start with the log (display log) and check if there is any IRF or interface event.
then check the stability of the switching part :
"display arp " and collect the mac address for 10.10.14.13 (do it a couple of times, and if mac-add changes, thent it's a IP conflict)
"display mac-addr <mac-addr collected> " multiple times to check that it persists and is seen on the same interface.
Post these traces here if needed.
------------------------------
Frederic MEUNIER
Original Message:
Sent: Sep 11, 2023 04:57 AM
From: spgsitsupport
Subject: Device attached to IRF member losing network
So I got this setup "working"
5900 JC772A IRF via HPE X140 40G QSFP+ MPO SR4 Transceivers
https://community.arubanetworks.com/discussion/5900-jc772a-irf-via-hpe-x140-40g-qsfp-mpo-sr4-transceivers
But devices attached to the Member 3 (one in another building) seem to drop off the network quite regularly in odd way (they will be there one moment, gone next & back again)
C:\Users\admin>ping sw-gardenbuildingPinging sw-gardenbuilding [10.10.14.13] with 32 bytes of data:Reply from 10.10.14.13: bytes=32 time=5ms TTL=254Request timed out.Request timed out.Request timed out.Ping statistics for 10.10.14.13: Packets: Sent = 4, Received = 1, Lost = 3 (75% loss),Approximate round trip times in milli-seconds: Minimum = 5ms, Maximum = 5ms, Average = 5msC:\Users\admin>ping sw-gardenbuildingPinging sw-gardenbuilding [10.10.14.13] with 32 bytes of data:Reply from 10.10.14.13: bytes=32 time=1ms TTL=254Reply from 10.10.14.13: bytes=32 time<1ms TTL=254Reply from 10.10.14.13: bytes=32 time<1ms TTL=254Reply from 10.10.14.13: bytes=32 time<1ms TTL=254Ping statistics for 10.10.14.13: Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),Approximate round trip times in milli-seconds: Minimum = 0ms, Maximum = 1ms, Average = 0ms
Where could I try to start troubleshooting?
Thanks
Seb
------------------------------
spgsitsupport
------------------------------