Cluster:Netgear GBIC issues: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
| Line 1: | Line 1: | ||
The symptoms | == The symptoms == | ||
Port trunking on the cairo Netgear GSM712 GBIC ports causes frequent, repeatable delays of a few seconds for no obvious reason. Delays on a single GBIC/pair of GBICs might also be repeatable. | Port trunking on the cairo Netgear GSM712 GBIC ports causes frequent, repeatable delays of a few seconds for no obvious reason. Delays on a single GBIC/pair of GBICs might also be repeatable. | ||
Things to remember | == Things to remember == | ||
* Is LACP stable? | * Is LACP stable? | ||
* Do we need LACP for port trunking? | |||
* Is STP/RSTP on Netgear products stable? | * Is STP/RSTP on Netgear products stable? | ||
* Could we try backing down to an earlier firmware? (We did a jump from 2.2.0→2.6.0 recently.) | * Could we try backing down to an earlier firmware? (We did a jump from 2.2.0→2.6.0 recently.) | ||
Things I've tried | == Things I've tried == | ||
* Auto LACP enabled/disabled | * Auto LACP enabled/disabled | ||
| Line 17: | Line 18: | ||
* Swapped bottom GBICs | * Swapped bottom GBICs | ||
* Bad port 12? | * Bad port 12? | ||
* Marked all GBIC modules with identifying scratches. | |||
== Tests == | |||
=== Single channel === | |||
* GBIC 1 in port 11 on SW0, GBIC 2 in port 11 on SW1, other GBICs depopulated: Pauses | |||
* GBIC 1 in port 11 on SW1, GBIC 2 in port 11 on SW0, other GBICs depopulated: Pauses | |||
Revision as of 18:54, 23 June 2005
The symptoms
Port trunking on the cairo Netgear GSM712 GBIC ports causes frequent, repeatable delays of a few seconds for no obvious reason. Delays on a single GBIC/pair of GBICs might also be repeatable.
Things to remember
- Is LACP stable?
- Do we need LACP for port trunking?
- Is STP/RSTP on Netgear products stable?
- Could we try backing down to an earlier firmware? (We did a jump from 2.2.0→2.6.0 recently.)
Things I've tried
- Auto LACP enabled/disabled
- STP/RSTP/No spanning-tree
- Swapping cables
- Swapped top GBIC
- Swapped bottom GBICs
- Bad port 12?
- Marked all GBIC modules with identifying scratches.
Tests
Single channel
- GBIC 1 in port 11 on SW0, GBIC 2 in port 11 on SW1, other GBICs depopulated: Pauses
- GBIC 1 in port 11 on SW1, GBIC 2 in port 11 on SW0, other GBICs depopulated: Pauses