Dale Ghent wrote: > On Oct 1, 2007, at 6:32 PM, John Forte wrote: > > >> The COMSTAR project team is beginning discussions around how >> COMSTAR will impact the user experience for the current iSCSI >> target. While we don't have all of the designs in place for the >> iSCSI target under COMSTAR, we have a feel for some items that will >> differ slightly from the existing iSCSI target implementation. >> > > John, excellent write up. I also echo Ben's thoughts with one > addition - statistics. > > I see COMSTAR being an excellent aggregator/provider for IO > statistics, status, and management, a la the IP stack for the various > TCP/IP-related SNMP MIBs. In this case, a consumer could implement > RFC 4544. > > Have there been any internal thoughts on this subject? > > /dale > _______________________________________________ > storage-discuss mailing list > [email protected] > http://mail.opensolaris.org/mailman/listinfo/storage-discuss > > Dale,
Statistics, on the whole, has been briefly discussed but we haven't reached any firm commitments here. However in terms of your mention of the iSCSI MIB and RFC 4544, I believe that would be something, if implemented, exclusive to the iSCSI target provider (and/or any iSCSI related modules) rather than COMSTAR itself (which is the framework only). Specifics of the provider transport (iSCSI, fibre channel, etc) will be, for all intents and purposes, private to the provider. So in that respect, it is only SCSI statistics that will be available via COMSTAR. However, that doesn't mean that more complete statistics cannot be modeled by separate module that has a view of all the components in the target side data path, i.e. the logical unit provider, the COMSTAR framework and the target port providers. - John _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
