Initial filesystem design synopsis.
dennis.melentyev at gmail.com
Thu Feb 22 03:16:29 PST 2007
2007/2/22, Matthew Dillon <dillon at apollo.backplane.com>:
:On Wed, 21 Feb 2007, Matthew Dillon wrote:
:> - Infinite snapshots
:> - Multi-master operation
:> - Infinite logless Replication
:transid space: monotonic increasing on each replication target, or a
:fine-grained synchronised timestamp*, or something else?
Monotonic increasing AND a fine-grained timestamp. Low bits of
the timestamp (sub-nanosecond equivalent) would simply be used to
Monotonic increasing transaction ids are *CRITICAL* to replication
protocols. Absolutely critical. It's the difference between having
to keep a physical log of changes (with unbounded size), and just
having to store the last transaction ID you had synchronized to.
Well, it does not match monolitic increasing requirement, but how
about having some "Master", distributing transaction ID's blocks of
numbers on segment's demand?
E.g. something like IPv4 IP blocks with self-maintained last bits?
Or, borrow some IPv6 ideas?
PS. If stupid, just throw this idea away.
More information about the Kernel