On Mar 1, 2010, at 11:51 AM, gonzalo diethelm wrote:
z = new ZMQ ();
s = z.makeSocket (ZMQ::PUB);
z = new ZMQ::Context ();
s = new ZMQ::Socket (z, ZMQ::PUB);
My preference is for the latter. Two reasons:
1. It's the current way it is. No need to change the API.
2. It's more obvious
On Mar 1, 2010, at 12:17 PM, Martin Sustrik wrote:
Hi Chuck,
For the Java and Ruby cases, why not move to a factory pattern where
a class method instantiates and returns an object containing the
context and all of the other bits?
// Java z = ZMQ.factory(); s = z.makeSocket(ZMQ::PUB);
#
On Apr 15, 2010, at 11:18 AM, Martin Lucina wrote:
All,
as part of some API and naming cleanups (more to core) that we'd like to do
before the 2.0.7 release I would like to solicit comments for/against
renaming the ZMQ_P2P socket type to ZMQ_PAIR.
Rationale: The name ZMQ_P2P evokes a)
On Apr 21, 2010, at 8:49 AM, Pieter Hintjens wrote:
On Wed, Apr 21, 2010 at 3:33 PM, Martin Sustrik sust...@imatix.com wrote:
Ruby is basically a single OS thread + green Ruby threads, right?
Language != implementation of language.
See here
I took a look at the Java code and saw references to P2P so I know it hasn't
been updated to use the new PAIR constant. That leads me to believe it needs a
particular build of the C libraries. If I want to experiment with the Java
bindings, which C release does it match up with?
Thanks...
cr
I'm going to generate some FFI bindings for running this C lib under jruby. Is
master stable enough for me to do this work or should I use the 2.0.6 release?
As along as I'm asking, is the 2.07 release imminent?
cr
___
zeromq-dev mailing list
I am trying to install zmq on OSX. I successfully built and installed the
library. However, when I tried to build the ruby bindings, I got an error.
http://gist.github.com/391694
I'm usually pretty good at figuring this stuff out, but that error tells me
*nothing* useful. I thought that
I installed a newer patch release of 1.8.7 (p249) and it now compiles. The perf
tests work too.
Sorry for the noise.
cr
On May 5, 2010, at 9:01 PM, Chuck Remes wrote:
I am trying to install zmq on OSX. I successfully built and installed the
library. However, when I tried to build the ruby
, Chuck Remes wrote:
I installed a newer patch release of 1.8.7 (p249) and it now compiles. The
perf tests work too.
Sorry for the noise.
cr
On May 5, 2010, at 9:01 PM, Chuck Remes wrote:
I am trying to install zmq on OSX. I successfully built and installed the
library. However, when I
Both zmq_bind and zmq_connect have a paragraph similar to this in their docs:
A single socket may be connected to an arbitrary number of peer addresses
using
_zmq_connect()_, while also having an arbitrary number of local addresses
assigned to it using _zmq_bind()_.
I admit I don't
On May 15, 2010, at 2:49 PM, Pieter Hintjens wrote:
On Sat, May 15, 2010 at 9:32 PM, Chuck Remes cremes.devl...@mac.com wrote:
I admit I don't understand how a socket can be a parameter to multiple bind
or connect calls. Perhaps this makes sense to folks who have been using
sockets
I have a question about zmq_msg_t managment when using send with the
ZMQ_NOBLOCK flag.
If the send fails, is it proper to call zmq_msg_close on the current zmq_msg_t?
With the ruby bindings, I don't call zmq_msg_close until the message goes out
of scope and the garbage collector collects it.
I just wrote this to a friend of mine who is looking at 0mq. It took me a while
to understand 0mq (not sure I really do yet) but his reaction to it left me
thinking that this might be useful to someone else.
Note that I realize that bind connect aren't *always* flip-flopped but with
0mq I
On May 24, 2010, at 8:07 AM, Martin Sustrik wrote:
5. Real men prefer to be in control rather than being controlled.
Therefore they don't like callbacks. For quiche-eaters it's easy to
create a wrapper that would translate received messages into callbacks.
Now that is just mean. :)
cr
I'd like to announce the availability of a new Ruby reactor library using 0mq.
ZMQMachine is patterned after the popular Ruby Eventmachine library which
itself borrowed heavily from the Python Twisted framework.
http://github.com/chuckremes/zmqmachine
It's rather immature but I am already
Guys,
thank you for all of your hard work. 0mq is a very cool and very worthy
project. I'm glad to be involved with it even in my own small way.
cr
On Jun 4, 2010, at 1:24 PM, Pieter Hintjens wrote:
Nice work, Mato!
-Pieter
Sent from my Android mobile phone.
On Jun 4, 2010 7:56 PM,
I'd like to request some docs for the 0mq devices (queue, forwarder and
streamer). Looking at the C they are quite simple but probably deceptively so.
For example, what mix of sockets is allowed for each device? I don't see any
code that throws errors if, for example, the queue device has a
FYI, the ruby FFI bindings have been updated to match what is available in the
2.0.7 release. Find them here:
http://github.com/chuckremes/ffi-rzmq
And the ruby reactor library has also been updated to use the API changes.
http://github.com/chuckremes/zmqmachine
cr
On Jun 7, 2010, at 6:34 AM, Martin Lucina wrote:
I would also prefer this solution. Martin, what is wrong with keeping the
current state of affairs and, at your option, either creating an
experimental or socket-intergration branch for your work on closer
integration with the BSD socket API,
On Jun 7, 2010, at 1:29 PM, Matt Weinstein wrote:
Folks -
I'm developing on Mac OS/X (Leopard), and deploying on CENTOS.
configure --with-pgm fails PGM fails with:
checking if the PGM extension is supported on this platform...
configure: error: the PGM extension is not
On Jun 11, 2010, at 10:28 AM, Matt Weinstein wrote:
Folks,
In re: patterns UPSTREAM and DOWNSTREAM are a bit ambiguous, e.g. some folks
think the packets will go DOWNSTREAM vs. UPSTREAM. I know we're back to
the verb/noun/adverb/whatever argument, but if you have to think about it...
On Jun 11, 2010, at 1:05 PM, Martin Sustrik wrote:
One problem that results is that the receiving socket would be in the
ready state after receiving the response, but the rest of the
network would not reflect that. If another request were sent with the
same identity it could trounce
On Jun 30, 2010, at 9:23 AM, Derek Developer wrote:
Is there any way to break the blocking of zmq_recv() without firing up a
client and sending a message like QUIT SERVER?
No. The only way to deal with this situation right now is to do a non-blocking
receive and use zmq_poll to test for
On Jul 8, 2010, at 6:55 AM, Matt Weinstein wrote:
IMHO the least surprising close semantics depend on the class of socket.
For example, REQ/REP or PAIR should not bother to flush, because they
are a communicating pair (temporal/ry or not), and if the
correspondent has gone away, that's
On Jul 11, 2010, at 1:54 AM, Ryan Chan wrote:
Hello,
I have a customized crawler project which consists of 4 subsystems.
a. Links discovery (5+ EC2 instance)
b. Web crawlers (20+ EC2 instance)
c. Links DB (MongoDB x 1)
d. Contents DB (MongoDB x 1)
1. Web crawlers (b) fetch links in
On Jul 12, 2010, at 1:04 PM, Peter Alexander wrote:
On Mon, Jul 12, 2010 at 1:45 PM, Chuck Remes cremes.devl...@mac.com wrote:
On Jul 12, 2010, at 12:18 PM, Steven Clark wrote:
Forgive me a simple/stupid question here-
To make use of ZeroMQ, do both ends of a simple network connection
On Jul 12, 2010, at 1:26 PM, Peter Alexander wrote:
Chuck you know more about this stuff then I do, but if he wanted to,
couldn't he use a standard socket in the 0mq application to
communicate with the with the foreign device. Granted zmq_poll()
does add a little more overhead.
On Jul 16, 2010, at 12:34 PM, Dmitri Toubelis wrote:
I have a scenario when I need to use PUB/SUB scheme with multiple publishers.
Is it possible to do with inproc transport?
Here is what I tried:
- create PUB socket and bind it to inproc://pipe endpoint
- create another PUB socket and
On Jul 16, 2010, at 1:06 PM, Dmitri Toubelis wrote:
I call bind only once. The second time I use connect and it returns 0.
Well, I clearly misread the original message. If that's the case, then I don't
know what the problem is. Hopefully someone more familiar with the codebase can
chime in.
On Jul 19, 2010, at 5:40 AM, Peter Alexander wrote:
Do any of you have the history of irc log files?
I'm going to put up for public hosting and search-able.
I saw on irc that you mentioned you have a hole in your archives. I *might*
be able to fill it since I've been lurking on that
On Jul 29, 2010, at 10:50 PM, Eric Bell wrote:
What is the behavior when subscriber in a publish-subcribe pattern changes
the message topic (aka message type) it is subscribed to?
I ran a test where I changed the message topic and then repeatedly read all
messages from the socket until
I have a client (XREQ socket) that connects to a server (XREP socket) via an
intermediary Forwarder device proxy (XREP XREQ bound to different ports).
I can successfully send data from the client to the server. I see the traffic
flow through each step by printing the message contents inside
On Aug 25, 2010, at 2:39 PM, Chuck Remes wrote:
On Aug 25, 2010, at 2:16 PM, Chuck Remes wrote:
I have a client (XREQ socket) that connects to a server (XREP socket) via an
intermediary Forwarder device proxy (XREP XREQ bound to different ports).
I can successfully send data from
On Aug 25, 2010, at 2:57 PM, Chuck Remes wrote:
On Aug 25, 2010, at 2:39 PM, Chuck Remes wrote:
On Aug 25, 2010, at 2:16 PM, Chuck Remes wrote:
I have a client (XREQ socket) that connects to a server (XREP socket) via
an intermediary Forwarder device proxy (XREP XREQ bound
I opened a ticket for a problem I have noticed since 2.0.7 that persists in
2.0.8.
http://www.greenpasture.org/community/?q=node/290
I am wondering if I am misusing the framework. Briefly, the crash occurs when
calling zmq_close from another thread on a socket that is blocked on send/recv.
It
On Aug 30, 2010, at 1:31 PM, Wolfgang Richter wrote:
I don't think it should automatically append a byte. The guarantee that 0MQ
sends exactly the bytes I give it is perfect. Modifying data almost
invisibly isn't good in my opinion.
Maybe I just like the simplicity 0MQ gives me.
I
On Aug 31, 2010, at 8:15 AM, Mikael Helbo Kjær wrote:
Hi everyone
I've been tinkering with my system and I've hit a problem inside zmq that I
haven't been able to debug or figure out just yet. My code is trying to send
when I hit it.
I get an access violation (for those who aren't
On Aug 31, 2010, at 11:04 AM, Pieter Hintjens wrote:
Martin,
That is a brilliant idea.
I'd like to likewise organize a meetup in Chicago on September 19th,
since I'll be there that weekend.
-Pieter
On Tue, Aug 31, 2010 at 3:10 PM, Martin Sustrik sust...@250bpm.com wrote:
Hi all,
On Aug 31, 2010, at 11:23 AM, Pieter Hintjens wrote:
On Tue, Aug 31, 2010 at 6:12 PM, Chuck Remes cremes.devl...@mac.com wrote:
I'm based in Chicago and would like to hook up with you and any other local
0mq folks.
Great!
Sunday 19th, maybe you can suggest a good place where the beer
On Aug 31, 2010, at 12:57 PM, John McLaughlin - PTIR wrote:
I have a need that does not seem to have a natural solution in 0MQ, but maybe
I am missing something.
Here’re the basics:
There will be a Server process (with a “published” address)
There will be Client processes that may
On Sep 3, 2010, at 12:43 AM, kasicass wrote:
I want to use 0MQ to build a network infrastructure for online game. But
the REQ/REP, PUB/SUB pattern is not suitable for it.
My requirement is
1. N client_proc connect to 1 server_proc
2. server_proc need to know one client's enter and
On Sep 3, 2010, at 7:06 AM, Pieter Hintjens wrote:
On Fri, Sep 3, 2010 at 1:31 PM, Martin Sustrik sust...@250bpm.com wrote:
I don't think there's a way to accommodate both existing install base
and new apps written with stable branch (currently 2.0.8). I hate to say
it, but the fix will
On Sep 3, 2010, at 10:33 AM, Martin Sustrik wrote:
Hi Brian, Chuck, Lestrrat,
There's eintr branch in zeromq/zeromq2 repo at github. The blocking
calls (zmq_send, zmq_recv and zmq_poll) should return EINTR in case of
Ctrl+C. Can you please check it whether it works as expected?
After
I have to do a rolling upgrade across my servers. Is it safe to have a mix of
1.6 and 1.4 servers in a master/slave configuration where the master is 1.6.2?
cr
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
Sorry, wrong list!
cr
On Sep 6, 2010, at 1:38 PM, Chuck Remes wrote:
I have to do a rolling upgrade across my servers. Is it safe to have a mix of
1.6 and 1.4 servers in a master/slave configuration where the master is 1.6.2?
cr
___
zeromq
On Sep 11, 2010, at 5:49 AM, Pieter Hintjens wrote:
Hi all
To those in or near the Chicago area, we're doing a meetup at Orso's
Italian restaurant at 6pm on Sunday 19th.
Let me know if you want to come so we can book enough places.
Save me a seat.
cr
On Sep 13, 2010, at 1:00 PM, Stefan Majer wrote:
Hi,
I´m implementing a application which is basically a kind of log
collection framework. It is basically divided into two parts:
- message producers, many of them are installed on a node and
produces log messages, send via zmq to 2 message
Right now the zmq and ffi-rzmq gems have slightly different APIs. In a perfect
world, an implementation detail like C extension versus FFI would be completely
hidden so users could switch back and forth without worrying about code
breakage.
I did a quick audit of the zmq gem versus the
On Sep 24, 2010, at 1:41 PM, Praveen Baratam wrote:
I am trying to write a pub-sub server using ZMQ.
The simplest way to describe my application is its a messaging server to
which clients connect using normal tcp sockets over intranet/internet. The
purpose is to exchange messages between
On Sep 24, 2010, at 10:35 PM, Praveen Baratam wrote:
Dear Chuck,
Thanks for writing back.
I did take a look into the zmq_reactor. the examples were not so clear and
there is no documentation of the library and its api. i guess i should infer
everything from the code.
I was
I am wondering if the master branch is far enough along that bindings
maintainers should start work on adopting the new APIs. Guidance, please.
cr
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
On Oct 15, 2010, at 12:43 PM, David Wolfe wrote:
Hi, all--
Can anyone offer advice on how to reduce zeromq's memory usage for an
embedded app? I think zmq_init() uses the default stack size (8192 KB
in Linux), which is a pretty big chunk when you've only got 64MB to work
with(!) Does
On Oct 19, 2010, at 1:21 PM, Pierre Menard wrote:
Does anyone know if it would be possible to get zeromq compiled for an
fpga and has anyone done this? Any thoughts or hints greatly
appreciated.
I would bet that most people trying to use 0mq on FPGA are using it for
trading. That means it
I think you are comparing apples and oranges.
SCTP is an actual transport protocol.
0mq is a socket library that can be layered on top of any kind of transport.
Right now it supports inproc (comm between threads), ipc (interprocess
communication), pgm (multicast), and tcp (you know what this
On Nov 2, 2010, at 3:31 PM, Chris Wong wrote:
Hi,
Hi,
I'm starting a project to migrate existing messaging code to ZeroMQ. The
new zmq_close/term() semantic in 2.1 is important for this project. Do we
already have a rough roadmap on approximately when 2.1 will be out of beta?
There is an issue opened to track this problem:
https://github.com/zeromq/zeromq2/issues#issue/66
Does anyone with more C/C++ experience than I have a way of solving this issue?
A lot of Ruby folks use the Ruby binaries from rubyinstaller.org which includes
an optional mingw toolchain for
I was running a program that allocated a couple of hundred REQ sockets to
connect to a server with a single REP socket via a QUEUE device.
I noticed that when I allocated 64 sockets I started getting an assertion
failure.
Assertion failed: nbytes == sizeof (command_t) (mailbox.cpp:193)
I
On Nov 7, 2010, at 11:14 PM, Chuck Remes wrote:
I was running a program that allocated a couple of hundred REQ sockets to
connect to a server with a single REP socket via a QUEUE device.
I noticed that when I allocated 64 sockets I started getting an assertion
failure.
Assertion
On Nov 8, 2010, at 5:35 AM, Martin Lucina wrote:
Chuck, what system are you on? Linux?
Mato, this is on OSX 10.6.4 Intel.
You guys figured out it was a non-blocking send which I didn't mention in my
original post, so good detective work there. The message size is approximately
2kbytes.
A lot of testing and discussion regarding this issue occurred on irc. Here's a
summary.
1. OSX reports very strange values when the 0mq library runs getsockopt for
SO_SNDBUF. It returns 32. The thought is that it is returning kbytes instead of
bytes like every other platform. :(
2. The
On Nov 8, 2010, at 1:12 PM, Chuck Remes wrote:
A lot of testing and discussion regarding this issue occurred on irc. Here's
a summary.
1. OSX reports very strange values when the 0mq library runs getsockopt for
SO_SNDBUF. It returns 32. The thought is that it is returning kbytes instead
On Nov 15, 2010, at 9:27 AM, Martin Sustrik wrote:
On 11/15/2010 02:18 PM, Chuck Remes wrote:
I discussed this issue with Janak on IRC. When he says that the
publisher stops he means that the process goes down. Two minutes
later he restarts it but the subscribers do not see any further
On Nov 15, 2010, at 2:30 PM, Mikko Koppanen wrote:
Hello,
short background (and correct me if something is wrong here):
The sys://log transport is an internal transport, a PUB socket
providing logging information on what happens inside 0MQ. The
information has not yet been defined but
On Nov 18, 2010, at 11:00 AM, T-zex wrote:
Thank you Pieter,
We are using 2.0.10, and trying to create a forwarder which forwards
messages from multiple threads to multiple subscribers. The forwarder
listens to multiple inproc publishers. This forwarder also implements
tcp subscriber
On Nov 18, 2010, at 11:51 AM, T-zex wrote:
What we have is:
inproc(a)---
|
Incoming(c)
|
inproc(b)--
And the question is when we bind a and b and then do connect on c, is
message delivery guaranteed form a and b to c? or do we need to do
On Nov 18, 2010, at 3:22 PM, Praveen Baratam wrote:
Hello All!
I have a particular situation where client nodes connect to server nodes to
solve tasks (PUSH-PULL). At any given point there are many clients (PUSH -
Connect) connected to all the available servers (PULL - Bound) and each
On Nov 22, 2010, at 1:33 PM, Scott Asher wrote:
Hello,
Just querying about expecting behavior in 2.1.0. If I'm connecting a
ZMQ_PAIR and I bind() one end and then send() before I connect the other end,
my send() blocks, even though I've set HWM on the socket to 0, and I've got
On Nov 23, 2010, at 9:31 AM, Martin Sustrik wrote:
Hi Koert,
Having used zeromq in a few small applications I now am used to the
behavior that it typically doesn’t matter if a connect happens before a
bind. It is great. However how does it work?
If say I have a server on machine 101 and
I'm unsure how zmq_term() is supposed to behave in 2.1.0 and later. The docs
describe its blocking behavior like so:
After iterrupting all blocking calls, _zmq_term()_ blocks until all sockets
open within the context are closed using _zmq_close()_ and either all the
pending outbound
On Nov 27, 2010, at 3:01 PM, Chuck Remes wrote:
[snip]
Is zmq_term() supposed to call zmq_close() on all known sockets on my behalf
or do I need to keep a list of these sockets and iterate over them calling
zmq_close()?
It causes odd behavior when running under Ruby. I cannot interrupt
On Nov 27, 2010, at 3:12 PM, Chuck Remes wrote:
On Nov 27, 2010, at 3:01 PM, Chuck Remes wrote:
[snip]
Is zmq_term() supposed to call zmq_close() on all known sockets on my behalf
or do I need to keep a list of these sockets and iterate over them calling
zmq_close()?
It causes odd
On Nov 27, 2010, at 3:15 PM, Chuck Remes wrote:
https://github.com/zeromq/zeromq2/issues/issue/127
I'm talking to myself here... :)
This blocking behavior was raised as an issue in #85.
https://github.com/zeromq/zeromq2/issues#issue/85
That ticket hasn't been updated, but the discussion
On Nov 27, 2010, at 3:31 PM, Martin Sustrik wrote:
On 11/27/2010 10:24 PM, Chuck Remes wrote:
This blocking behavior was raised as an issue in #85.
https://github.com/zeromq/zeromq2/issues#issue/85
That ticket hasn't been updated, but the discussion did include references
to SO_LINGER
On Nov 28, 2010, at 10:27 AM, Pieter Hintjens wrote:
On Sat, Nov 27, 2010 at 10:53 PM, Chuck Remes cremes.devl...@mac.com wrote:
Wait a second...
Chuck, there is a long thread (search for 'issue 85') which tries to
dissect this problem. Martin likes to answer with puzzles like
On Dec 1, 2010, at 2:57 PM, Praveen Baratam wrote:
Hello All!
I was just reading the Advanced Stuff chapter of the guide and found the XREP
socket as a router pretty interesting.
I just need a clarification that is not found anywhere in the guide or else
where.
XREP(Router) uses
On Dec 1, 2010, at 4:19 PM, Praveen Baratam wrote:
Dear Chuck,
Thanks for the quick reply.
Well if ever an Identity is reused, I guess the earlier socket connected with
that identity will be orphaned and all the messages will be routed to the new
socket with that ID. Effectively the
On Dec 2, 2010, at 2:41 PM, Jay Banyer wrote:
Gday,
I'm new to 0MQ and my first attempt at using it is stuck. zmq_connect() for a
SUB socket is failing with errno == 22.
Code is here: http://pastebin.com/HctDf64J
Compile with: g++ -l zmq -o zmq_test file.cpp
Run publisher with:
On Dec 7, 2010, at 9:40 AM, Scott Asher wrote:
I've run into this assertion occasionally when I close sockets. Questions
about both parts:
1) !more
Why do you care if there is part of a multi-part message still left on the
pipe enough to assert about it? Maybe this should be a
On Dec 9, 2010, at 1:08 AM, Mahadevan R wrote:
On Thu, Dec 9, 2010 at 10:52 AM, Mahadevan R mdeva...@gmail.com wrote:
On Wed, Dec 8, 2010 at 3:44 PM, Mahadevan R mdeva...@gmail.com wrote:
Hi guys,
I'm getting a strange segfault with the following stack:
#0 0x7f623e64edd0 in
On Jan 10, 2011, at 6:26 PM, David Kantowitz wrote:
Hello,
In v2.1 zmq_term() will block if there are any open sockets -- in previous
versions of ZMQ zmq_term() was non-blocking.
For my application this causes problems, so I've changed my copy of
ctx::terminate() to be non-blocking. I
On Jan 18, 2011, at 1:14 PM, Michael Kogan wrote:
Hello,
I am new to ZeroMQ, and I am trying to prototype a messaging bus for our Java
application.
However, I am encountering an issue with the test classes ZMQQueueTest.java
and ZMQForwarderTest.java - they hang.
Running the debugger,
On Jan 18, 2011, at 2:40 PM, Michael Kogan wrote:
Thank you Chuck.
Never occurred to me the unit tests are wrong. Adding .close() calls made the
unit tests complete successfully.
They are wrong as of the release of 2.1.0 (the behavior of zmq_term() changed).
I don't play with Java much
On Jan 19, 2011, at 10:14 PM, Seth Burleigh wrote:
Im having the same problems as the link below. I would like to close a socket
- but while im closing it, it is possible that other messages are already in
my sockets queue. Thus, when i close it, messages are lost. ideally, a
socket
On Feb 2, 2011, at 3:33 AM, Andrew Hume wrote:
i don't have permissions to tcpdump.
i'm running 2.0.9
Unless you are running your application on ports less than 1025, I don't think
you need elevated privileges to attach to the ports with tcpdump. You should be
able to run it as a regular
All of the articles that used to be available on the tutorials page from the
old site (before the reorg a few months back) are orphaned. They can be found
if you search the site for the correct terms, but there is no table of
contents page where all of them can be found.
cr
Due to some ongoing issues with 0mq on OSX, I switched over to using my linux
box as the main dev and test server.
I am running a very recent master from the last day or two, so it's all 2.1.0.
My systems do a lot of high-volume communication amongst 4 distributed
components. They connect
On Feb 15, 2011, at 6:00 PM, Chuck Remes wrote:
Due to some ongoing issues with 0mq on OSX, I switched over to using my linux
box as the main dev and test server.
I am running a very recent master from the last day or two, so it's all 2.1.0.
My systems do a lot of high-volume
On Feb 15, 2011, at 7:41 PM, Dhammika Pathirana wrote:
Assertion failed: new_sndbuf old_sndbuf (mailbox.cpp:183)
new_sndbuf = 10485760, old_sndbuf = 10485760
Is it failing on last check?
new_sndbuf = 4 * old_sndbuf, is that correct?
Right, the assertion is verifying that new_sndbuf
On Feb 16, 2011, at 6:54 AM, Ian Barber wrote:
On Wed, Feb 16, 2011 at 2:30 AM, Chuck Remes cremes.devl...@mac.com wrote:
So, it's trying to expand the buffer past 10MB and fails. The buffer starts
out at 512k (new default I set) and grows very rapidly. What kind of data is
0mq putting
On Feb 16, 2011, at 9:08 AM, Pieter Hintjens wrote:
On Wed, Feb 16, 2011 at 3:47 PM, Chuck Remes cremes.devl...@mac.com wrote:
1. The publisher pushed too much data to its socket...
I fixed it by finally detecting and repairing the weekend clock skip bug.
So it's not really fixed in the 0mq
On Feb 16, 2011, at 10:00 AM, Pieter Hintjens wrote:
Hi Chuck,
Here's a step-by-step explanation in pure 0mq terms to replicate...
OK, here's those steps translated into raw C: https://gist.github.com/829599
It works on a variety of Linux boxes here, on 0MQ/2.1.0 as well as the
latest
I have a reproducible case and a full explanation here:
https://github.com/zeromq/zeromq2/issues#issue/165
It turns out it was unrelated to the PUB/SUB configuration. Instead, an
unrelated piece of code was creating two REP sockets with the same identity.
When any communications were initiated
On Feb 16, 2011, at 2:33 PM, Michael Kogan wrote:
The code is very simple with a single INPROC forwarder within process and a
TCP Forwarder to talk to the other node. I am also adding ZooKeeper to
discover and connect and configure the nodes. I am getting CPU utilization
that is just
On Feb 17, 2011, at 7:33 AM, Martin Sustrik wrote:
On 02/17/2011 02:08 PM, Pieter Hintjens wrote:
Yes, that fixes it and it's now impossible, as far as I can tell, to
break a server this way.I've a client that reconnects very rapidly,
and a server that sleeps, and the mailbox does not
On Feb 19, 2011, at 4:22 PM, Pieter Hintjens wrote:
On Sat, Feb 19, 2011 at 7:42 PM, Pieter Hintjens p...@imatix.com wrote:
On Sat, Feb 19, 2011 at 7:08 PM, Martin Sustrik sust...@250bpm.com wrote:
Wow! Great that wikidot allows for auto-uploading now.
It's still in beta but works pretty
On Feb 22, 2011, at 4:18 PM, Yusuf Simonson wrote:
Hello,
I posted earlier about Efficient and reliable pub/sub mechanism. Thanks for
the replies, they set me off in the right direction.
While developing my application, I have noticed a couple issues with jzmq's
performance and was
On Feb 22, 2011, at 5:45 PM, Yusuf Simonson wrote:
So do you see why the addition of a sleep call doesn't make sense? That
is, how does a call to sleep make a message magically arrive?
I haven't tested enough to be 100% sure, but I think what sleep is doing is
giving the application
On Feb 23, 2011, at 11:19 AM, Michael Kogan wrote:
Martin, I have been thinking about this, but is it possible to identify gaps
if I am subscribing to channels? What about multiple publishers.
Michael,
what is a channel? That isn't a 0mq concept as far as I know.
If you have multiple
On Feb 23, 2011, at 11:31 AM, Pieter Hintjens wrote:
On Wed, Feb 23, 2011 at 6:25 PM, Chuck Remes cremes.devl...@mac.com wrote:
If you really need reliable delivery of each published message, then the
Pub/Sub pattern is not appropriate. You should use the REQ/REP pattern
because you'll
On Feb 25, 2011, at 2:50 PM, Michael Kogan wrote:
Would it be a good idea for the Socket.unregister method to take an index as
well a a socket? Seems like we are wasting time running through an array?
I suggest keeping the API as simple as possible. If you need a faster search,
hide those
1 - 100 of 381 matches
Mail list logo