Talk:Transaction processing

I disagree with the implicit assertion in the article that transaction processing has moved away from DBMSs. Perhaps a discussion/link of distributed transactions is relevant? Clausen 00:24 15 Jun 2003 (UTC)

article only mentions double-copy implementation

The article says that transactions are implemented using the double copy method. This is incomplete and misleading. At least some mention should be made of journaling methods as a means of implementing transactions, or the sentences describing the double copy method should be turned into a link to another article describing transaction implmentations.

Chris Connett 18:47, 7 October 2005 (UTC)

ACID and this article

This article and the ACID article need to be merged, as they cover a great deal of the same ground. I've just largely rewritten this article; I'll be happy to merge the two at a later date when I have a bit more time to work on it. The article also needs some expansion. Agateller 19:26, 27 August 2006 (UTC)

Disambiguation helpful....

What term might I look up to find info on electronic financial transaction processing? Thanks. --Remi0o 11:06, 11 March 2007 (UTC)

Merge proposal

I think Transaction Processing System should be merged into Transaction processing. There is much overlap, and System issues can become a section within the more general article. (SEWilco 06:11, 9 April 2007 (UTC))

Transaction Processing and a susyem by the same name do not belong in the same article because they are not the same type of entity. To suggest that they do belong together implies that concepts such as "Batch Processing" and "Batch Processing Systems" belong together. These things are fundamentally different in that both Transaction Processing" and Batch Processing" are comparable and each has at its core, the mathematical basis (contraints) that makes each unique. A system -- any system -- is a model that represents a method of implementing processes. An implementation is not necessarily the definition of the process that is implemented.

That combining these different abstractions may appear reasonable is likely due to incorrectly partitioning certain attributes of each. For example, the acronym ACID appears in each. Under "Transaction processing" the subheading for "ACID" contains, in place of relevant content, subjective ideas with no relevence to the fact that "ACID" is an acronym. (e.g. to say that 'it shifts the time of job processing to when the computing resources are less busy' is a criterion that applies to batch processing, which is by definition, different from transaction processing.) Under "Transaction Processsing System" the subheading for "ACID" contains a definition for each term in the acronym but fails to state that it is an acronym. (That this may be self-evident is subjective, not objective, and so does need to be stated).

Agree (sort of). See my proposal below. The articles should be discussing different things, but there is so much overlap it's hard to disentangle the threads. If no one else takes it on I'll hack at it, but I'd rather not try it myself -- or at all, if someone else competent wants to try. It think what's involved is moving chunks of text from Transaction Processing System to this article and deleting the redundancies. I added a HAT note to this article to this effect.Peter Flass (talk) 13:17, 14 November 2012 (UTC)

Transaction Isolation Levels?

Would this be a good place to discuss this concept? (I am using Microsoft jargon here, I don't know what lingo is useful in discussion of Oracle or other products). It mainly has to do with locking and what is visible inside the transaction. I think it certainly is noteworthy. Is there another article that discusses this? Root⁴(one) 19:38, 20 November 2007 (UTC)

See Isolation (database systems). - JCLately (talk) 04:53, 21 November 2007 (UTC)

Thanks. BTW, Is a reason you or anybody can think of not to include a link to that here (in article)? I would do it myself, but since you did provide me the link... I dunno. It just sounds to me like a good idea to me to include it and I'm wondering why it is not already there! I just don't want to step on anybody's toes. Root⁴(one) 05:49, 23 November 2007 (UTC)

There is no explicit mention of Isolation in this article, so I don't think it makes sense to add that link without considering the bigger picture. For example note that the ACID section of this page is totally lame and really doesn't cover ACID at all. This and many other pages under the heading of Category:Transaction processing could use some serious editing. You are most welcome to attempt any improvements if you are so inclined. Don't worry about stepping on anyone's toes - no one owns this material. Feel free to use the Talk page when in doubt, but don't be too shy about making changes directly if they seem like obvious, non-controversial improvements. If someone disagrees or is inspired to further improve your edits, he or she can make further changes without anyone's taking it personally. - JCLately (talk) 15:13, 23 November 2007 (UTC)

Merge Extreme Transaction Processing

I am proposing that we merge Extreme Transaction Processing into the Implementation section article, at least until that technology becomes notable enough to warrant a separate article. Comments/Objections? Joshua Scott (talk) 04:37, 11 January 2010 (UTC)

I would be in favor of such a merger. We need a source for the definition of the term, too. DavidLeeLambert (talk) 22:48, 6 August 2011 (UTC)

I am for deletion - this sounds like an arbitrary number and a marketing term, not anything actually meaningful to industry. Given this request has been open two years, I am proposing deletion on that page now. prat (talk) 01:09, 17 May 2012 (UTC)

Agree, either merge or delete Extreme Transaction Processing. See other merger propsal below. Peter Flass (talk) 13:07, 13 November 2012 (UTC)

• oppose Extreme Transaction Processing (and the more closely-related event processing) are distinct from transaction processing. It's recognised as a different discipline, with its own products and literature. Transaction processing has been around commercially since the '60s and the advent of telecommunications that could separate data clients from a processing centre (i.e. computer access moved out of the computer room itself, and was no longer demanded in a single, simple batch) XTP has arisen in the last decade or so, to address the volume and synchronisation problems of really big data volumes and transaction rates. Different approaches are needed to cope with these. Andy Dingley (talk) 09:11, 12 August 2013 (UTC)

• Note that the article with capital letters seems to be lifted mostly from a book by that name (where the capital letters would be explained). It seems to have been a popular expression in 2011, but I would say redirect that one here. Event processing is indeed different, since it often relaxes the ACID requirements etc. Six years later, about time to do something. W Nowicki (talk) 23:33, 18 March 2017 (UTC)

Transaction manager and Transaction processing monitor

Transaction manager and Transaction processing monitor both redirect to this article. What do these terms mean? --Abdull (talk) 17:03, 8 May 2012 (UTC)

Merger proposal

I have proposed that Transaction processing system be merged into this article, in whole or in part, because there seems to be considerable overlap. I would suggest that Transaction processing system might be worthwhile to retain solely as a discussion of specific systems: UNIVAC TIP, IBM CICS, XDS CP-V transaction processing, whatever Burroughs had, DEC's systems for PDP-10 and VAX, Tuxedo, etc. Note that this article has more on that topic than the other. Peter Flass (talk) 13:04, 13 November 2012 (UTC)

OK, guess not. Peter Flass (talk) 01:05, 5 March 2015 (UTC)

Implementations

What about CICS and ACP/TPF? These would seem to be the first things to come to mind when discussing transaction processing. Peter Flass (talk) 13:12, 14 November 2012 (UTC)