Date: Thu, 14 Jul 2022 15:37:46 +0200 From: Roger Pau =?utf-8?B?TW9ubsOp?= <roger.pau@citrix.com> To: Brian Buhrow <buhrow@nfbcal.org> Cc: freebsd-xen@freebsd.org Subject: Re: How deep is the xl socket queue? Message-ID: <YtAcKtKZ%2BVUj/BlS@MacBook-Air-de-Roger.local> In-Reply-To: <202207140322.26E3M9Wr025120@nfbcal.org> References: <202207140322.26E3M9Wr025120@nfbcal.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jul 13, 2022 at 08:22:09PM -0700, Brian Buhrow wrote: > hello. Recently, I've started seeing messages like the following: > sonewconn: pcb 0xfffff801074d7800 (local:/var/run/xenstored/socket): Listen queue overflow: 2 > alread > y in queue awaiting acceptance (1 occurrences) > > I figured out what causes this. I have a number of automated scripts that use the xl > command to do their work. Sometimes, multiple scripts run at once, causing too many instances > of xl to run simultaneously. My questions are: > What is the default queue depth for the xl socket listener? Hm, I've got no idea TBH. Note however this is the xenstored socket, libxl instances are the clients here. Can you check which version of xenstored are you currently running? Just paste the output of: # ps aux |grep xenstored Does it make any difference if you use Ocaml or C xenstored implementations? You can do that by changing the line in /usr/local/etc/rc.d/xencommons: XENSTORED=/usr/local/sbin/xenstored Or: XENSTORED=/usr/local/sbin/oxenstored And rebooting (note: you must reboot the system in order to switch between xenstored implementations). The binary prefixed with 'o' is the Ocaml one, while the binary without any prefix is the C one. Thanks, Roger.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?YtAcKtKZ%2BVUj/BlS>