Hello,
I'm trying to figure out concurrent database transactions, because I
realized that I don't fully understand transactions. I wanted to find out
if there are any negative implications of a transaction being open for a
long time (do you maybe know if there are?). So I created the following
snippet:
require "sequel"
require "logger"
DB = Sequel.postgres("transactions")
DB.create_table!(:users) do
primary_key :id
String :age
end
id = DB[:users].insert({age: 24})
DB.logger = Logger.new(STDOUT)
pid1 = fork {
DB.transaction do
sleep 2
end
}
sleep 1
pid2 = fork {
DB.transaction do
DB[:users].update(age: 25)
end
}
Process.waitpid pid2
puts DB[:users].first
Process.waitpid pid1
What I wanted to do is that the transaction from the second subprocess is
"nested" inside the first (that it happens between BEGIN...COMMIT of the
first one), because I wanted to produce some abnormal behaviour (not
important). What I would expect, since Postgres should support concurrent
transactions, that everything goes fine. But it goes boom (gist
<https://gist.github.com/janko-m/94ab65e246aa483021b0>). Why do I get the
warning that there is already a transaction in progress, if Postgres
supports concurrent transactions?
Sorry if this is not Sequel-related, but I really want to figure out what's
going on :)
Thanks,
Janko
--
You received this message because you are subscribed to the Google Groups
"sequel-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/sequel-talk.
For more options, visit https://groups.google.com/d/optout.
