Hello,
- sys.maxint too
It might be bad design to have long running transactions, but there
are cases you can't avoid that.
E.g. evolving a generation on a huge DB.
Saturday, October 17, 2009, 10:30:43 AM, you wrote:
LR 2009/10/16 Lennart Regebro rege...@gmail.com:
So HTTP seems to contradict
2009/10/16 Lennart Regebro rege...@gmail.com:
So HTTP seems to contradict itself, typically But from looking at
other peoples responses, most interpret the specification that the
connection should immediately be closed, so the Zope does the right
thing there, and Varnish should be changed.
On Sat, Oct 17, 2009 at 10:30, Lennart Regebro rege...@gmail.com wrote:
2009/10/16 Lennart Regebro rege...@gmail.com:
So HTTP seems to contradict itself, typically But from looking at
other peoples responses, most interpret the specification that the
connection should immediately be closed, so
Hi again,
Izak Burger wrote:
I eventually distilled the check varnish uses into a small C program,
and an interesting problem shows up. When you call shutdown(fd, SHUT_WR)
on your socket connection, in effect telling zope that you're done
talking to it, it looks like zope responds in kind
2009/10/16 Izak Burger i...@upfrontsystems.co.za:
Secondly, since zope still requires older python
versions to run
Zope 2.12 run on Python 2.6. 2.6.4a1 is coming out the weeken, so if
this is a Python bug and we want it in 2.6.4 we need to get it in
before a b1, which also probably will be
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Izak Burger wrote:
Hi again,
Izak Burger wrote:
I eventually distilled the check varnish uses into a small C program,
and an interesting problem shows up. When you call shutdown(fd, SHUT_WR)
on your socket connection, in effect telling zope
On Fri, Oct 16, 2009 at 16:36, Tres Seaver tsea...@palladion.com wrote:
[...]
You might also look at fixing varnish: I don't know of any valid
reason for it to be using the half-open connection model to test that
an HTTP-based backend is up
Going out on a limb here, but I think Varnish might
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Leonardo Rochael Almeida wrote:
On Fri, Oct 16, 2009 at 16:36, Tres Seaver tsea...@palladion.com wrote:
[...]
You might also look at fixing varnish: I don't know of any valid
reason for it to be using the half-open connection model to test that
2009/10/16 Tres Seaver tsea...@palladion.com:
Hmmm? A TCP socket corrresponds to exactly one open file descriptor,
which has to stick around for the response data to come back on.
Half-closing it is just silly, and is only guaranteed to work where
both ends expect to handle this case.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Lennart Regebro wrote:
2009/10/16 Tres Seaver tsea...@palladion.com:
Hmmm? A TCP socket corrresponds to exactly one open file descriptor,
which has to stick around for the response data to come back on.
Half-closing it is just silly, and is only
Tres Seaver wrote:
You might also look at fixing varnish: I don't know of any valid
reason for it to be using the half-open connection model to test that
an HTTP-based backend is up -- certainly no browser in the world does
that; instead, modern browsers nearly always try to keep the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Izak Burger wrote:
Tres Seaver wrote:
You might also look at fixing varnish: I don't know of any valid
reason for it to be using the half-open connection model to test that
an HTTP-based backend is up -- certainly no browser in the world does
Leonardo Rochael Almeida wrote:
Going out on a limb here, but I think Varnish might be trying to save
on file-descriptors.
Interestingly, The Squid FAQ almost seems to imply that closing the
write-side can cause more file-descriptors to be used. Squid can
apparently not tell the difference
From the HTTP/1.1 spec:
Servers SHOULD NOT close a connection in the
middle of transmitting a response, unless a network or client failure
is suspected.
But on the other hand:
When a client or server wishes to time-out it SHOULD issue a graceful
close on the transport connection.
14 matches
Mail list logo
