Subject | Re: [Firebird-Architect] FB2, read-commited and blobs |
---|---|
Author | Ann W. Harrison |
Post date | 2005-11-30T20:22:42Z |
Vlad Horsun wrote:
collection of record versions that contain blobs. In fact, record
versions with blobs can be garbage collected if a newer version of the
record shares the same blob id. So, it's not exactly parallel.
Regards,
Ann
>Not exactly. Read-committed transactions only block the garbage
>>3) Create a new lock series for blob garbage collection - oldest_blob
>>for example.
>
>
> As i understand this new lock series have the same meaning for
> read-comited transactions as old lock series for snapshot transactions ?
collection of record versions that contain blobs. In fact, record
versions with blobs can be garbage collected if a newer version of the
record shares the same blob id. So, it's not exactly parallel.
> If yes - may be it is better to name it using 'read-commited' words ?Again, not exactly parallel.
> Are we have symmetry here with snapshot\read-commited transactions
> and new\old lock series ?
>Snapshot transactions won't take this lock at all.
>
>>4) At startup, read-committed transactions also take a lock in the blob
>>garbage collection series, identified by transaction id and containing
>>the current oldest-active as the data portion.
>
>
> What value snapshot transactions will put in this lock ? Zero ? Or its
> will not take this lock at all ?
> If latter - why read-commited transactionsBecause we need that lock to wait on when there's an update conflict.
> take usual transaction lock ?
Regards,
Ann