jimklimov

• Linux /etc/init.d/virtualbox-hdd-timeout:

#!/bin/sh

### BEGIN INIT INFO
# Provides:          hdd-timeout
# Required-Start:    
# Required-Stop:     
# Should-Start:      
# Should-Stop: 

jimklimov

Earlier "icc" emulated "gcc"; nowadays it is a "clang"

* wget from intel site https://software.intel.com/content/www/us/en/develop/tools/oneapi/base-toolkit/download.html?operatingsystem=linux
* run the shell installer (may need to `export TERM=vt100` in console mode)
* use wrapper script below as /usr/bin/icc (symlinks as i++ and icc-cpp, maybe suffixed with version from path component e.g. icc-2021.3.0)
----
# copy-paste the file as /usr/bin/icc
chmod +x /usr/bin/icc
( cd /usr/bin && for I in i++ icc-cpp icc-cpp-2021.3.0 icc-2021.3.0 i++-2021.3.0 ; do ln -s icc $I ; done )
----

jimklimov

sed -e 's,.\[8mha://.*.\[0m,,' < log > %D/1089-log

jimklimov

= Self-wrapping code()

I have my own version of "declarative" we use, and in that case you can add parameters like timestamps: true so I have code roughly (see below) like:

def workflow = { // stuff to do }
if (config.timestamps){
    //*******NEEDED *****//
    def inner = workflow()
    //********************//
 workflow = { timestamps { inner() } }

jimklimov

Installing prereqs on OpenIndiana for maximum build

Achievable setup and result:
./configure --with-all --with-dev --with-doc=auto --without-modbus
gmake all -j8
gmake check

Prereqs:

pkg install gnu-make automake autoconf \

jimklimov

omnios pxe install notes

The goal was to be able to perform a network install of OmniOS from a Debian system.

I know there is refinement that can take place in this process (and these notes), but this got me up and running.

All testing took place on a Mac running VirtualBox with the extra extensions installed to allow for PXE booting. However I cannot see why this wouldn't work on real hardware in a network that is already setup to do PXE installs.

I setup:

jimklimov

Add the following chunk to your existing ISC dhcpd.conf file.

if exists user-class and ( option user-class = "iPXE" ) {
    filename "http://boot.smidsrod.lan/boot.ipxe";
}
else {
    filename "undionly.kpxe";
}

(or see https://gist.github.com/4008017 for a more elaborate setup

jimklimov

# User ssh configuration file ~/.ssh/config
# Gist https://gist.github.com/terrywang/3997931
# man ssh_config for more information
# Inspired by the blog post below to fight the NSA
# https://stribika.github.io/2015/01/04/secure-secure-shell.html

# Outside of the firewall, with HTTPS proxy
# Package connect-proxy is required.

# Amazon EC2

jimklimov

# Add-on logic for OBS server config to allow only certain hosts to build certain projects

# For specific build constraints, for example to require a specific
# security level of the workers.
# Return 0 (false) if this worker is not suitable for this job.
# Return 1 (true) to allow execution of this job on this worker.
# Used from bs_dispatch via
#   { next if !$BSConfig::dispatch_constraint->($ic->{$job}, $workerinfo{$idle}); }
#

jimklimov

(2020-08-10 02:40:30) jimklimov: cheers all who won't sleep :) doing some ZFS replications and hit that annoyance again about how slow it is to `zfs recv` many small snapshots. Is there some tunable to speed it up a bit? Looking at speed numbers in the console, it seems that for every incremental snapshot it spends over 60s regardless of size. So empty 312byte snapshots take a minute at average 5 b/s, and larger multimegabyte snaps take 70s, and huge ones average at relatively decent looking speeds in MB/s range. What can I do to not wait that 1 minute every snap?
(02:47:40) jimklimov: looking at the consoles, I see it doing zero I/O on source and destination pools (zpool iostat 1), then the send|recv pipe says "receiving incremental stream of X into Y" and there is a burst of I/O for a second, and the pipe logs how it received a few kilobytes in a lot of seconds...
(02:58:43) jimklimov: using Adam's scripts from http://dtrace.org/blogs/ahl/2014/08/31/openzfs-tuning/ I see that the destination pool txg-syncin