Subject | Re: Ignorance or ancient bug? |
---|---|
Author | alex_vnru |
Post date | 2002-01-18T14:13:35Z |
--- In ib-support@y..., "Pavel Cisar" <pcisar@u...> wrote:
this kind at FB site, to dissipate widespread mistake:
JO> Q1 will return the number of records that existed at the start of
JO> the query
JO> process. Anything added after the start would not be included.
JO> IB will grab
JO> transaction numbers used for determining the correct records at
JO> the beginning of
JO> the query. Q2 has not committed its data at that time so its
JO> transaction ID
JO> will not be considered.
I quote Jeff as English-speaking man to don't add more confusion
to this already confused subject by my barbaric English.
Best regards, Alexander V.Nevsky.
> Transactions have to be _always_ isolated (ACID) by definition, butThank you, Pavel. Perhaps it can be placed in FAQ or something of
> we're going with compromise to get better throughput (avoid strict
> serialization). Any aggregation function returns "accurate" results
> only in true isolation from other transactions, for FB in snapshot
> isolation levels.
>
> So, it's not IMHO bug or mistake in implementation, it's how world
> works :-)
this kind at FB site, to dissipate widespread mistake:
JO> Q1 will return the number of records that existed at the start of
JO> the query
JO> process. Anything added after the start would not be included.
JO> IB will grab
JO> transaction numbers used for determining the correct records at
JO> the beginning of
JO> the query. Q2 has not committed its data at that time so its
JO> transaction ID
JO> will not be considered.
I quote Jeff as English-speaking man to don't add more confusion
to this already confused subject by my barbaric English.
Best regards, Alexander V.Nevsky.