Hello,
I have strange problem of dropping link connectivity with our CRS317-1G-16S+.
Description of setup:
We have HP switch stack with 10Gbit stacking link with several (approx. 4) free 10Gbit ports.
We needed more 10Gbit ports for our graphics workstations(windows) and one disk array(linux)
so
we decided to buy one CRS317-1G-16S+ connect it to HP with 2x 10Gbit LACP agreggation with trunked VLANs
and
connect workstations and array to CRS.
We use Mikrotik SFP+ modules for 10Gbit copper ethernet with CAT6A cables.
Problem:
in undefined intervals windows workstations are dropping out from network - unpredicatbly.
Windows show link 10Gbit connected, non-zero packets sent, zero received.
CRS shows link 10Gbit connected, sending packets occassionaly, zero receiving.
Beavior observed on several ,cables’’ cables were tested with cable meter - they comply with CAT6A.
Observed on Intel LANs and on Edimax LANs. Edimax was replaced by Intel.
Known workaround:
remove SFP+ module from CRS, wait 10 secs and plug in again.
This will make work it again - for a day, sometimes for a week.
Configuration of CRS:
/interface ethernet
set [ find default-name=sfp-sfpplus1 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus2 ] l2mtu=10218 loop-protect=off mac-address=B8:69:F4:E0:4E:47 mtu=10194
set [ find default-name=sfp-sfpplus3 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus4 ] l2mtu=10218 loop-protect=off mac-address=B8:69:F4:E0:4E:49 mtu=10194
set [ find default-name=sfp-sfpplus5 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus6 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus7 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus8 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus9 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus10 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus11 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus12 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus13 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus14 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus15 ] l2mtu=10218 loop-protect=off mtu=10194
set [ find default-name=sfp-sfpplus16 ] l2mtu=10218 loop-protect=off mtu=10194
/interface bridge
add admin-mac=B8:69:F4:E0:4E:57 auto-mac=no name=bridge0 protocol-mode=none vlan-filtering=yes
/interface bonding
add mode=802.3ad mtu=10194 name=sfp-bond0 slaves=sfp-sfpplus1,sfp-sfpplus2
add mode=802.3ad mtu=10194 name=sfp-bond1 slaves=sfp-sfpplus3,sfp-sfpplus4 transmit-hash-policy=layer-2-and-3
/interface bridge port
add bridge=bridge0 frame-types=admit-only-vlan-tagged ingress-filtering=yes interface=sfp-bond0
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-bond1 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus12 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus13 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus14 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus15 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus16 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus10 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus11 pvid=201
add bridge=bridge0 frame-types=admit-only-untagged-and-priority-tagged ingress-filtering=yes interface=sfp-sfpplus9 pvid=201
/interface bridge vlan
add bridge=bridge0 tagged=sfp-bond0 vlan-ids=201
where bond0 is ,uplink’’ to HP stack with LACP and bond1 is connection to disk array(linux).
to stack there is LACP with ,trunk’’ with tagged VLAN
to diskarray there is LACP without tagging
to workstations there is simple access port
Could anyone review my configuration?
Could anyone tell me, what can cause aforementioned behavior? (i.e. dropping of computers from the network)
This is our live production environment, so reliability is quite important.