Rich Adamson wrote: .
Last, the bonding of two nics at the server level _requires_ the associated switch interface to support the exact same bonding algorithm. Historically, that has been a problem for many switch vendors.
Not so sure I understand, but if you mean, 'the algorithm to select a link(usually a hashing of some layer source/destination to ensure sequence)' ? This would not really to match in both ends, AFAIK.
In some implementations I've worked with, the non-use of proprietary protocols to 'establish/maintain' the LAG/whateveriscalledit group would force you to use 'static' assignments of the members, but other than that, not big deal..(counting on RFI for failure detect and etc helps...)
Short answer... I'd never do it. Long answer... think in terms of high availability "systems"; the nic card is the least concerning.
There are quite few carriers using similar 'bonding' to dual core ethernet routing switches doing "split-MLT", where the 2 chassis would 'look like' a single box with bonding/FEC/MLT links, so is more than NIC card only. I would go as far as say that most of the time is done for redundancy, not for bandwidth (call signaling and announcements only voice bearer)
Anyway, it would seem to me the original poster was looking for redundancy, not really 'added bandwidth'.
_______________________________________________ --Bandwidth and Colocation provided by Easynews.com -- Asterisk-Users mailing list To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
