Peter Memishian wrote: > > There's more to it. > > > > Some other Sun products make use of the Solaris DHCP code. For example, > > the Sun Ray code depends heavily on it. > > ... and moreover, last I checked, unbundled products such as the Netra HA > suite have their own custom DHCP server datastore written to the > publically documented plugin API.
Unbundled Netra HA... ok, so how prevalent is the usage of Netra HA these days? I can't say. > There are some surprising things included in the justification for > replacing the server, such as: > > | 3) Scalability and Performance > | > | ISC DHCP daemon being a more modern code base scales and performs better > | than Solaris' out-of-date DHCP daemon even when ISC is using a text file > | for IP information vs. Solaris' SUNWbin format. When the Solaris DHCP > | daemon is configured to use SUNWtext format, performance and scalability > | is even more noticeable. > > First, I presume "SUNWtext" is supposed to be "SUNWfiles". Second, I'd > like to understand how ISC is outperforming SUNWbinfiles, since when we > did this work (back in 2000), the Sun DHCP server was orders of magnitude > faster. We have tools like DTrace, so where's the analysis? Correct, SUNWfiles... In a recent test reported by a customer, ISC was approximately 20% in their medium-sized environment (350 class C subnets, each with approximately 240 IP addresses being served up) against the Solaris DHCP server using SUNWbin as the format. When using SUNWfiles (eg. text), it was at least 50% faster in most tests and some cases 2x. ISC has progressed significantly since 2000 while the Solaris daemon has stagnated. Again some of this is annecdotal, seeing what support or issues there are for or against replacing the current DHCP daemon. Certainly to point out that the "emperor has no clothes" for some points as well. Dave _______________________________________________ networking-discuss mailing list [email protected]
