« first day (1894 days earlier)      last day (3269 days later) » 

Paul White
Paul White
02:50
@billinkc Bill you have altogether too much spare time.
@Phil No. Some hardware items do affect plan selection e.g. #processors, amount of memory but there's no CPU/IO calibration as such.
 
4 hours later…
Mahesh
Mahesh
06:28
@swasheck Thought so. :)
@AndriyM It's actually weird. I could not get where .. in construct to work, but when I instead used where column = any($1::varchar[]) and passed in a list, the driver converted it properly to {item,item2,..} and passed it to postgres and it worked.
I suspect the where in issue is a bug in the specific driver I'm using, I see no reason why it shouldn't work. I'm no postgres expert but I have a hunch. :P
It was actually a colleague's bug, I don't do sql these days. We were using postgrex, the elixir postgres connector.
Andriy M
Andriy M
07:26
@Mahesh Ahh, I see. I was playing with where ... in on SQL Fiddle and couldn't get it to work either. Although that's perfectly understandable, as I am no Postgres person at all. Was really curious, though. Thanks.
 
3 hours later…
Michael Green
Michael Green
10:27
@MDCCL I suppose the Rs do disappear in cliched, Crocodile Dundee Australian, or 'Strine to some. With about one quarter of the population born overseas it is difficult to say just want "typical" is. One friend definitely pronounces beer as "bee-yah". But then she speaks Greek at home!
 
1 hour later…
ypercubeᵀᴹ
ypercubeᵀᴹ
11:42
@MichaelGreen ha! Beer in Greek sounds something like that indeed. Do they do the same in other similar words, like peer or dear that the greek word is totally different?
Tom V
Tom V
12:30
@ypercubeᵀᴹ Birra if I recall correctly?
ypercubeᵀᴹ
ypercubeᵀᴹ
Yes
Where "bi" sounds like be, not like bicycle.
Tom V
Tom V
12:46
@ypercubeᵀᴹ I may not know a lot, but I know how to order my beers abroad :)
3
Tom V
Tom V
12:56
@Phil Who else but @PaulWhite has a post on that sqlblog.com/blogs/paul_white/archive/2010/09/01/… (cc @MaxVernon)
@swasheck Ah thanks, After googling trace flags as database options, I didn't know that but it looks very nice
 
2 hours later…
Valentin Tihomirov
Valentin Tihomirov
15:16
How is it possible that distributed transactions are always made right despite your reads my be inconsistent?
 
2 hours later…
ypercubeᵀᴹ
ypercubeᵀᴹ
17:26
@ValentinTihomirov he explains it right there. because his model assumes that writes are done centrally, in one primary server. The reads are done from multiple replicas.
MDCCL
MDCCL
17:49
Could the following question be considered a shopping-list one?
0
Q: Does exist some software for transforming data from one DB schema to another?

Peter KoncI need to transform between two different database schemas the same data. For example i have one table in old schema like this: and i need to transform it to new schema like this: So i have few questions to ask: Does exist software that can migrate data between these two schemas (Without ...

database-design replication schema data-migration database-migration
Seems valid, but I'm not that sure.
@MichaelGreen So it's a particularly diverse country. Yes, it's hard to say which are the typical customs, then; perhaps the typical accent could be defined like that, diverse.
 
3 hours later…
ConcernedOfTunbridgeWells
ConcernedOfTunbridgeWells
20:43
@ValentinTihomirov He's talking about a hard ACID-style consistency model as opposed to the quorum replication found in systems like MongoDB. On a distributed database (or transaction) with ACID semantics, any replicas of the data will be updated before the transaction is committed.
Quorum replication means only a subset of the replicas will be updated immediately, and reads from other replicas may see old versions of the data until the lazy update process updates the replicas. This is a performance optimisation known as 'eventual consistency' used on some platforms such as MongoDB.
However, it can lead to applications reading obsolete versions of the data, so it may not be acceptable. In this case you need a platform with ACID semantics.
See dba.stackexchange.com/questions/34892/… for a more in-depth discussion.
 
2 hours later…
Andriy M
Andriy M
22:31
I'm voting to close this question as off-topic because it has been cross-posted at DBA.SE and has already got an answer there. Please do not cross-post. — Andriy M 16 secs ago

« first day (1894 days earlier)      last day (3269 days later) »