search for: uncorking

...100644 --- a/server/protocol.c +++ b/server/protocol.c @@ -398,6 +398,11 @@ send_simple_reply (struct connection *conn, reply.handle = handle; reply.error = htobe32 (nbd_errno (error, false)); + if (conn->cork) { + r = conn->cork (conn, true); + assert (r == 0); /* For now, only uncorking can fail */ + } + r = conn->send (conn, &reply, sizeof reply); if (r == -1) { nbdkit_error ("write reply: %s: %m", name_of_nbd_cmd (cmd)); @@ -413,6 +418,14 @@ send_simple_reply (struct connection *conn, } } + if (conn->cork) { + r = conn->cork (conn,...

Slightly RFC, as I need more time to investigate why Unix sockets appeared to degrade with this patch. But as TCP sockets (over loopback to localhost) and TLS sessions (regardless of underlying Unix or TCP) both showed improvements, this looks like a worthwhile series. Eric Blake (2): server: Add support for corking server: Cork around grouped transmission send()s server/internal.h | 3

In the recent commit 3842a080 to add SEND_MORE support, I blindly implemented the tls code as: if (SEND_MORE) { cork send } else { send uncork } because it showed improvements for my test case of aio-parallel-load from libnbd. But that test sticks to 64k I/O requests. But with further investigation, I've learned that even though gnutls corking works great for smaller

I am writing some R code that generates some figures. I am primarily developing on GNU/Linux, but will ultimately run the code on Windows in order to get WMF files I can integrate in an existing document. The win.metafile() function, of course, doesn't exist in the GNU/Linux version of R. I am contemplating writing an abstraction layer that runs the right thing, but I need to know how the

I am generating some graphs (on GNU/Linux using R 1.5.x) for a PowerPoint presentation my boss is giving next week. He just tried cut/pasted a plot off of a PDF file I had created, but he complained that the fonts were too small and fuzzy, and that he couldn't change the background. I have been playing with png(bg=transparent), but I am not sure what appropriate height,width parameters and

On Fri, Mar 24, 2023 at 11:47:20AM +0100, Florian Westphal wrote: > qemu-nbd doesn't set TCP_NODELAY on the tcp socket. > > Kernel waits for more data and avoids transmission of small packets. > Without TLS this is barely noticeable, but with TLS this really shows. > > Booting a VM via qemu-nbd on localhost (with tls) takes more than > 2 minutes on my system. tcpdump

I've got a data frame that looks like this: subject foo bar 2 1.7 3.2 2 2.3 4.1 3 7.6 2.3 3 7.1 3.3 3 7.3 2.3 3 7.4 1.3 5 6.2 6.1 5 3.4 6.9 ... That is, I've got multiple rows per subject. I need to compute summaries within categories where the subject has the same number

I am wondering if there are any apps affected by this bug (and perhaps keeping people from switching Unicorn). It''s a fairly esoteric case, so I probably won''t make another release until tomorrow (sleepy now, will probably screw something else up or realize something else is broken :) Anyways it''s pushed out to master and 1.1.x-stable in case people want^Wneed it

I am getting ready to load a bunch of data into R. The data is all numeric, but some of the numbers are integer codes representing non-numeric semantics. What is the best way to "fix" the data frames so that these compenents are recognized as "factors" or "ordered", as appropriate? Can I "assign" to some attribute of the data frame component, like the

Nu Mcafee titles released on Jun 06 20:29:48 MSK 2007 Adobe Creative Suite CS3 269$ Adobe Photoshop CS3 89$ Symantec Norton 360 29$ Microsoft Office 2007 79$ Microsoft Vista Business 79$ Nero 7 Premium 39$ Adobe Acrobat 8 Pro 79$ Adobe Flash CS3 Pro 59$ Windows XP Pro +SP2 49$ Adobe Premiere 2.O 59$ Macromedia Studio 8 99$ 0ffice2OO3 w/Contact Mgr 69$ Quickbooks

I've got a set of data that are weighted by a largish integer (ranging from about 50K to 1.5 million). I'd like to plot CDF's for this data taking into account the weighting. What do you recommend? Thanks! -- Russell Senior ``The two chiefs turned to each other. seniorr at aracnet.com Bellison uncorked a flood of horrible

This can hide bugs in Rack applications/middleware. Most other Rack handlers/servers seem to follow this route as well, so this helps ensure broken things will break loudly and more consistently across all Rack-enabled servers :) --- This will be in the next release, so fix your applications/middleware before hand... lib/unicorn/http_response.rb | 2 +- 1 files changed, 1 insertions(+), 1

Nu Mcafee titles released on Jun 06 20:29:48 MSK 2007 Adobe Creative Suite CS3 269$ Adobe Photoshop CS3 89$ Symantec Norton 360 29$ Microsoft Office 2007 79$ Microsoft Vista Business 79$ Nero 7 Premium 39$ Adobe Acrobat 8 Pro 79$ Adobe Flash CS3 Pro 59$ Windows XP Pro +SP2 49$ Adobe Premiere 2.O 59$ Macromedia Studio 8 99$ 0ffice2OO3 w/Contact Mgr 69$ Quickbooks

In the course of some work I have been doing for Revolution Analytics I have had the necessity of modifying the plm function so that it would not die halfway through fitting. In particular, I was able to more than halve the runtime (for my particular run) and improve its memory usage with three small modifications: 1.) Replacing throughout apply(X, 2, mean) with colMeans, similarly with

I'd like to generate some plots like you'd see on the old "normal probability graph paper", like the first plot in: <http://www.itl.nist.gov/div898/handbook/eda/section3/normprpl.htm> except the horizontal scale would have 1%, 5%, 25%, 50%, 75%, 95%, 99%, or similar quantiles, with associated tick/grid lines. [still hunting around for a good example...] something like

...Changes: - Fixed maildir syncing errors about wrong UIDs - Maildir was re-read all the time, even if we knew nothing had changed there - SSL/TLS connections used to have extra ~0.2s latencies because kernel was waiting for more data before sending it (at least with Linux). Fixed it by corking/uncorking before/after sending. I don't think the socket latencies are this bad normally though, I should probably look more into it if we're setting some socket flags badly (or does it happen specifically because we were using corking for the socket before SSL and kernel increased the timeouts becau...

> As a streaming server, it's fairly crucial for icecast to > send out data with as low a delay as possible (many clients > don't care, but some do). That's why we use TCP_NODELAY - we > actually WANT to send out data as soon as we can. Nagle is inherently unsuited for streams. NODELAY was (imho) ment for connections for which Nagle isn't sufficient and CORK is not

Here is the summary on predict when using an arima0 object: The arima0 object must be based on a time series vector. That is; x <- ts(xm1, frequency=12, start=c(1975,1)) x.ar <- arima0(x,order=c(1,1,1)) predict(x.ar,n.ahead=3) Thanks so much to Prof. Brian Ripley and David Brahm and other! Sincerely, Erin Hodgess

Any time we reply to NBD_CMD_READ or NBD_CMD_BLOCK_STATUS, we end up calling conn->send() more than once. Now that we've disabled Nagle's algorithm, this implies that we try harder to send the small header immediately, rather than batching it with the rest of the payload, which causes more overhead in the amount of actual network traffic. For interfaces that support corking (gnutls, or

Michael, With regard to your comment below: > As a streaming server, it's fairly crucial for icecast to send out > data with as low a delay as possible (many clients don't care, but > some do). That's why we use TCP_NODELAY - we actually WANT to send out > data as soon as we can. Can you explain how some clients depend on a low delay when receiving data from icecast? How