Re: Global snapshots

On 2020/09/10 18:01, tsunakawa(dot)takay(at)fujitsu(dot)com wrote:
> From: Fujii Masao <masao(dot)fujii(at)oss(dot)nttdata(dot)com>
>> But I'm concerned about that it's really hard to say there is no patent risk
>> around that. I'm not sure who can judge there is no patent risk,
>> in the community. Maybe no one? Anyway, I was thinking that Google Spanner,
>> YugabyteDB, etc use the global transaction approach based on the clock
>> similar to Clock-SI. Since I've never heard they have the patent issues,
>> I was just thinking Clock-SI doesn't have. No? This type of *guess* is not
>> safe, though...
>
> Hm, it may be difficult to be sure that the algorithm does not violate a patent. But it may not be difficult to know if the algorithm apparently violates a patent or is highly likely (for those who know Clock-SI well.) At least, Andrey-san seems to have felt that it needs careful study, so I guess he had some hunch.
>
> I understand this community is sensitive to patents. After the discussions at and after PGCon 2018, the community concluded that it won't accept patented technology. In the distant past, the community released Postgres 8.0 that contains an IBM's pending patent ARC, and removed it in 8.0.2. I wonder how could this could be detected, and how hard to cope with the patent issue. Bruce warned that we should be careful not to violate Greenplum's patents.
>
> E.25. Release 8.0.2
> https://www.postgresql.org/docs/8.0/release-8-0-2.html
> --------------------------------------------------
> New cache management algorithm 2Q replaces ARC (Tom)
> This was done to avoid a pending US patent on ARC. The 2Q code might be a few percentage points slower than ARC for some work loads. A better cache management algorithm will appear in 8.1.
> --------------------------------------------------
>
>
> I think I'll try to contact the people listed in Clock-SI paper and the Microsoft patent to ask about this.

Thanks!

> I'm going to have a late summer vacation next week, so this is my summer homework?
>
>
>> One alternative is to add only hooks into PostgreSQL core so that we can
>> implement the global transaction management outside. This idea was
>> discussed before as the title "eXtensible Transaction Manager API".
>
> Yeah, I read that discussion. And I remember Robert Haas and Postgres Pro people said it's not good...

But it may be worth revisiting this idea if we cannot avoid the patent issue.

Regards,

--
Fujii Masao
Advanced Computing Technology Center
Research and Development Headquarters
NTT DATA CORPORATION