similar to: [PATCH] Added NoDelay config option and nodelay subsystem option

The included patch for openssh 3.0.2p1 implements overlapping read requests for the sftp client. It should be able to handle weird cases such as shriking files and reordered responses. This is only the first shot, and I'd be happy for any comments. I plan to implement something similar for the write path if this works out well. The maximum number of outstanding requests is quite high at

On Tue, Jun 23, 2020 at 5:17 PM Aki Tuomi <aki.tuomi at open-xchange.com> wrote: > > > On 23/06/2020 12:15 Joe Wong <joewong99 at gmail.com> wrote: > > > > > > Hello, I have added 'y' as nodelay in my passdb query and I see the > following in log: > > > > dovecot: imap-login: Debug: Ignoring unknown passdb extra field: nodelay >

Hello, I have added 'y' as nodelay in my passdb query and I see the following in log: dovecot: imap-login: Debug: Ignoring unknown passdb extra field: nodelay does it mean this option is no longer supported in 2.3.10 ? - Joe -------------- next part -------------- An HTML attachment was scrubbed... URL:

In certain situations sftp download speed can be much less than that of scp. After many days of trying to find the cause finally I found it to be the tcp nagle algorithm, which if turned off with TCP_NODELAY eliminates the problem. Now I see it being discussed back in 2002, but it still unresolved in openssh-4.2 :( Simple solution would be to add a NoDelay option to ssh which sftp would set.

Hi All, I noticed a very weird phenomenon when I'm copying data (200KB image files) to our glusterfs storage. When I run only run client, it copies roughly 20 files per second and as soon as I start a second client on another machine, the copy rate of the first client immediately degrade to 5 files per second. When I stop the second client, the first client will immediately speed up

This time the problem is that the ssh server only sets TCP_NODELAY for interactive (tty) sessions or if X11 forwarding is enabled. Neither of which are true for the use of the sftp subsystem. This hurts upload performance for sftp/sshfs. I'm not sure why this hasn't cropped up earlier. Were there any TCP_NODELAY related changes in the sshd code recently? Is there a reason not to

In order to test my overlapping request path for sftp on another ssh server, I downloaded ssh2 version 3.1.0 from ssh.com. Having downloaded it, I decided to study the use of TCP_NODELAY in that implementation. Here's what I found: * Both ssh2 and sshd2 has a NoDelay config option which is false by default. * The ssh2 client does not enable or disable NoDelay because of a channel

I have a system in place where it appears that TCP will make a massive change in behavior mid-stream with existing SSH sessions. We noticed the issue first with an application using an SSH forward. However, we were able to rule that out by generating the same TCP characteristics by having a perl script dump text out to a terminal simulating a large data flow from the far end(ssh server) back

I have a distributed/replicated setup with Glusterfs 3.0.2, that I'm testing on 4 servers, each with access to /mnt/gluster (which consists of all directories /mnt/data01 - data24) on each server. I'm using configs I built from volgen, but every time I access a file (via an 'ls -l') for the first time, I get all of these messages in my logs on each server: [2010-05-04 10:50:30] W

On Sun, Feb 24, 2002 at 09:04:03AM +0100, Ricardo Galli wrote: > Sorry, didn't explain well. > > Nagle's algorithm (rfc896) buffers user data until there is no pending acks > or it can send a full segment (rfc1122). > > icecast doesn't need it at all, because it already sends large buffers and > the time to send the next buffers is relatively very long. IMO

> On 23/06/2020 12:19 Joe Wong <joewong99 at gmail.com> wrote: > > > > > On Tue, Jun 23, 2020 at 5:17 PM Aki Tuomi <aki.tuomi at open-xchange.com> wrote: > > > > > On 23/06/2020 12:15 Joe Wong <joewong99 at gmail.com> wrote: > > > > > > > > > Hello, I have added 'y' as nodelay in my passdb query and I

> On 23/06/2020 12:15 Joe Wong <joewong99 at gmail.com> wrote: > > > Hello, I have added 'y' as nodelay in my passdb query and I see the following in log: > > dovecot: imap-login: Debug: Ignoring unknown passdb extra field: nodelay > > does it mean this option is no longer supported in 2.3.10 ? > > - Joe > It is supported in 2.3.10. The extra

The patch below (against openssh 3.2.3p1) adds a CheckMaxStartups option, defaulting to yes, to determine whether sshd calls drop_connection(). The motivation behind this is twofold. In our environment, our timesharing machines get enough incoming connections that will trigger spuriously with the default value (10 forked unauthenticated connections) as well as some significantly higher values,

On 24/02/02 05:02, Jack Moffitt shaped the electrons to say: > > - The server didn't check for the status of the client's socket before > > the unblocking send(). This caused a disconnection at a minimun network > > congestion, causing a broken pipe error (Linux 2.4 behaviour?) in the > > network. I've just added a poll in sock.c.> > Can you send me this

This patch adds a new option to sshd, chroot_users. It has the effect of chroot()ing incoming ssh users to their home directory. Note: this option does not work if UsePrivilegeSeparation is enabled. Patch is based on OpenSSH 3.4p1. *** servconf.h@@\main\1 Tue Oct 1 17:25:32 2002 --- servconf.h Wed Oct 2 06:17:48 2002 *************** *** 131,136 **** --- 131,137 ---- char

http://bugzilla.mindrot.org/show_bug.cgi?id=594 Summary: SSH tries to set nodelay on non-sockets Product: Portable OpenSSH Version: -current Platform: All OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: ssh AssignedTo: openssh-bugs at mindrot.org ReportedBy: jimb at

> 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

Hi Denis, Thanks for help! After trying several things out I finally worked it out ;-) No I can connect with the useres from my WinDomain to the samba server - that's fine But: How do I create vald shares for the several groups? How can I set the rights for the different folders for different Windows-users & Windows-groups? Has this be done on windows or on linux? I just tried to set

Hello everybody, I'm a C/C++ consultant working for Ericsson. I changed the OpenSSH-Portable code to add a new criteria into the Match sshd_config option read by the sshd server. The new criteria is "Subsystem"; so a conditional block based on subsystem client request can now be added to the sshd_config configuration server file to override settings in its global section.

I added a few features to openssh for my local use that I think would be more broadly useful. I basically added access control lists to control who would be allowed public key authentication. I added four config file entries for the server: PubkeyAllowUsers PubkeyDenyUsers PubkeyAllowGroups PubkeyDenyGroups These follow the same sematics as the already existing entries for