Re: logging with iptables
>> On Thu, 19 Sep 2024, fxkl4...@protonmail.com wrote: > in my iptables i havetcp LOG flags 0 level 4 prefix "REJECT: " > this does what i want but how to direct the logging > it gets written to multiple file in /var/log > syslog, messages, kern, debug > can i restrict this to a single file You might have better luck if you used rsyslog. I've appended the rsyslog.conf file I used on my last Linux box. Features: * I still use the short date format in typical syslog files -- I don't need the full year because each logfile is linked to a dated version in a subdirectory. This also makes log rotation vastly simpler. See the "TEMPLATES" section: /var/log/cron -> /var/log//MMDD/cron etc. * You can weed out crap messages that fill up logs. See the "FILTERS" section for more. * The first entry in the "RULES" section shows how to put iptables stuff in its own "firewall" log. -- Karl Vogel I don't speak for anyone but myself Photographing a volcano is just about the most miserable thing you can do. --Robert B. Goodman, who has clearly never tried to use a PDP-10 # - # rsyslog v5 configuration file # # $Revision: 1.4 $ $Date: 2020-08-31 01:07:59-04 $ # $Source: /doc/sitelog/linuxwks/vcs/etc/rsyslog.conf,v $ # $Host: linuxwks $ # $UUID: 0aac27dd-1bdd-3c91-92e8-857140e878db $ # # For more information see /usr/share/doc/rsyslog-*/rsyslog_conf.html # Problems? http://www.rsyslog.com/doc/troubleshoot.html MODULES = # provides support for local system logging (e.g. via logger command) $ModLoad imuxsock # provides kernel logging support (previously done by rklogd) $ModLoad imklog # provides --MARK-- message capability #$ModLoad immark # Provides UDP syslog reception #$ModLoad imudp #$UDPServerRun 514 # Provides TCP syslog reception #$ModLoad imtcp #$InputTCPServerRun 514 # GLOBAL DIRECTIVES === # Use short timestamp format ## $ActionFileDefaultTemplate RSYSLOG_TraditionalFileFormat $ActionFileDefaultTemplate ShortForm # Use high-precision timestamps and timezone information. ## $ActionFileDefaultTemplate RSYSLOG_FileFormat # File syncing capability is disabled by default. This feature is # usually not required, not useful and an extreme performance hit ## $ActionFileEnableSync on # Include all config files in /etc/rsyslog.d/ $IncludeConfig /etc/rsyslog.d/*.conf TEMPLATES === $template DYNauth,"/var/log/%$YEAR%/%$MONTH%%$DAY%/secure" $template DYNcron,"/var/log/%$YEAR%/%$MONTH%%$DAY%/cron" $template DYNfirewall,"/var/log/%$YEAR%/%$MONTH%%$DAY%/firewall" $template DYNkern,"/var/log/%$YEAR%/%$MONTH%%$DAY%/kernlog" $template DYNmail,"/var/log/%$YEAR%/%$MONTH%%$DAY%/maillog" $template DYNmessages,"/var/log/%$YEAR%/%$MONTH%%$DAY%/messages" # This is identical to traditional format, without the hostname. $template ShortForm,"%timegenerated% %syslogtag%%msg%\n" FILTERS = # This apparently comes from Radeon fence code in the kernel: ignore. #:msg, contains, " armed on ring " ~ #:msg, contains, " signaled from irq context" ~ #:msg, contains, " pending" ~ #:msg, contains, "alloc_contig_range:" ~ RULES === # Log iptables drops to firewall log using discard action. if \ $syslogfacility-text == 'kern' \ and $msg contains 'Denied' \ then?DYNfirewall & ~ kern.* ?DYNkern # # Log all the mail messages in one place. Postfix stuff must be ID'd # by looking at the message contents; this has to come before general # processing so we can discard the message, or postfix entries are # also logged to "messages". if \ $msg contains 'postfix/' \ or \ $syslogfacility-text == 'mail' \ then?DYNmail & ~ # # Log anything of level info or higher. # Don't log private authentication messages! ## authpriv.none;auth.none;kern.none;cron.none;local5.none;*.info ## /var/log/messages if \ $syslogseverity <= '6' \ and ( \ $syslogfacility-text != 'auth' \ and \ $syslogfacility-text != 'authpriv' \ and \ $syslogfacility-text != 'cron' \ and \ $syslogfacility-text != 'kern' \ and \ $syslogfacility-text !=
Re: BASH reference for those who are "learning by doing"?
On Mon 09 Sep 2024 at 04:47:20 (-0400), Anssi Saari wrote: > debian-u...@howorth.org.uk writes: > > > As a mere bifocal (well vari-focal) wearer can I suggest a different > > approach. Stop wearing tri-focals or any other variable focus specs for > > reading a computer screen. Tell them to get a [very cheap] pair of > > single focus reading glasses made to suit the distance their screen is > > away. > > I don't agree at all. I've used fixed focus glasses before but I find > close range varifocals a huge upgrade. They're extremely useful for > monitor work *and also* I can see and read things around me that fall > outside the extremely narrow focus range of single focus reading > glasses. I wonder if fixed focus even works for modern large two or > three monitor setups? I suppose it's fine for a single small screen. Have you tried some different fonts? My eyesight is poor, and a good font made all the difference. https://bezoar.org/posts/2023/0214/font-screenshots/ -- Karl Vogel I don't speak for anyone but myself Debugging: when you're the detective, the victim, and the criminal. --seen on Reddit, 11 Jul 2024
Re: bash history
This is how I keep a long-term record of bash commands from different sessions: https://www.reddit.com/r/bash/comments/ak9c3r/ HTH -- Karl Vogel I don't speak for anyone but myself Comment: I use a screwdriver a lot Reply: I'm all out of orange juice. Will straight vodka work? --Reddit "What tool helped you as an early sysadmin?"
Re: Markup in mail messages
>> On Thu, May 16, 2024 at 09:48:23AM -0400, to...@tuxteam.de wrote: > Actually I've been tempted to teach my mail reader to transform HTML > into some lightweight markup (yeah, you need a bit of heuristics for > that ;-) -- say Org, but why not its poor sister Markdown. https://github.com/aaronsw/html2text/ might interest you. It converts (relatively) sane HTML into Markdown. I put html2text.py into $HOME/lib and use this to call it: #!/bin/sh #
Re: Zutty fonts - zutty always uses the same font and fontsize
On Wed, May 01, 2024 at 08:32:31AM -0400, Sirius wrote:
> If Debian still packages it, look for rxvt instead, or use xterm. Both
> are well tried and well tested for when you want something.. dated. ;)
I resemble that remark. Xterm v390 was released on 19 Feb 2024, and
building it from source is easy.
https://invisible-island.net/archives/xterm/xterm-390.tgz{,.asc}
--
Karl Vogel I don't speak for anyone but myself
My mind is like my browser: 19 open tabs, three of them are frozen, and
I have no clue where the music is coming from.
Re: Changing The PSI Definition
>> On Fri, Jan 26, 2024 at 07:42:30AM -0500, Dan Ritter wrote:
> Might be time for a new font. I like Inconsolata, but l1I!
> should never look similar, nor O0@ or S$.
My eyesight sucks like a black hole with daddy issues, so I like fonts
that are a bit larger than most. My favorites on xterm:
* xft:Menlo-Regular:pixelsize=20:bold
* xft:Bitstream Vera Sans Mono:pixelsize=21:bold
* xft:Cascadia:pixelsize=22:bold
* xft:FiraMono-Regular:pixelsize=22
Your examples above are very readable with Menlo. These aren't bad:
* xft:Edlo:pixelsize=21:bold
* xft:Inconsolata-Bold:pixelsize=25:bold
* xft:Meslo LG S:pixelsize=20:bold
* xft:Meslo LG S:pixelsize=21:bold
* xft:UbuntuMono-B:pixelsize=25:bold
I run two xterms side-by-side on a 23-inch monitor:
/usr/local/bin/xterm -geometry 80x40-0+0 -j -b 10 -sb \
-si -sk -ls -u8 -sl 4000 -cr blue -bd black -bg white \
-fa xft:Menlo-Regular:pixelsize=20:bold -title Remote
For browsing (Firefox), my "prefs.js" file holds:
user_pref("browser.display.use_document_fonts", 0);
user_pref("font.default.x-western", "sans-serif");
user_pref("font.internaluseonly.changed", false);
user_pref("font.minimum-size.x-western", 18);
user_pref("font.name.monospace.x-western", "DejaVu Sans");
user_pref("font.name.sans-serif.x-western", "sans-serif");
user_pref("font.name.serif.x-western", "DejaVu Serif");
user_pref("font.size.fixed.x-western", 18);
user_pref("font.size.variable.x-western", 18);
--
Karl Vogel I don't speak for anyone but myself
Slogan of 105.9, the classic rock radio station in Chicago:
"Of all the radio stations in Chicago...we're one of them."
Re: smartctl cannot access my storage, need syntax help
On Tue, Jan 23, 2024 at 06:05:29AM -0500, gene heskett wrote: > On 1/23/24 00:30, Karl Vogel wrote: > >>> On 1/22/24 11:31, gene heskett wrote: > > > > G> How does an 8T backup server sound for another $200 in hdwe? Very > > G> enticing and I do have the sheckel's. > > > > https://www.amazon.com/dp/B07CQJBSQL > > Seagate Desktop 8TB external Hard Drive, 3.5 Inch, USB 3.0 STGY8000400 > > $168.18 > > > > What if you buy two, use one for a complete backup and the other for > > incrementals or differentials? (I know, more than $200...) > > My disastrous experience with the last pair of seagates preclude exploring > that path, ever again. Sorry, the Seagate was just an example -- I prefer Western Digital myself. My only point was using one or two external drives for backups. -- Karl Vogel I don't speak for anyone but myself And as we all know from experiments conducted during the Korean War, Diane, sleep deprivation is a one-way ticket to temporary psychosis. --FBI Special Agent Dale Cooper, "Twin Peaks"
Re: smartctl cannot access my storage, need syntax help
>> On 1/22/24 11:31, gene heskett wrote: G> How does an 8T backup server sound for another $200 in hdwe? Very G> enticing and I do have the sheckel's. https://www.amazon.com/dp/B07CQJBSQL Seagate Desktop 8TB external Hard Drive, 3.5 Inch, USB 3.0 STGY8000400 $168.18 What if you buy two, use one for a complete backup and the other for incrementals or differentials? (I know, more than $200...) -- Karl Vogel I don't speak for anyone but myself I yam Popeye of Borg. You will be askimilgrated.
Re: Isolated Web Co Session crash Firefox-ESR
On Wed, Dec 06, 2023 at 06:04:36AM +0100, to...@tuxteam.de wrote:
> On Wed, Dec 06, 2023 at 02:42:32AM +0800, jeremy ardley wrote:
>
> > I have discovered a magic bullet for solving running out of memory
> > sudo sync; sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
> > Sadly it looks like I'll need to do this daily,
>
> It's the browsers eating your memory. That's what they do.
I've had problems with Firefox eating my swap on both Linux and FreeBSD.
My fix has been to run the swap2ram script below hourly.
--
Karl Vogel I don't speak for anyone but myself
Constipational carry.
--NY Post 28 Nov "Suspect found hiding handgun in his rectum" comment #4
# --
#!/bin/ksh
# /dev/null
case "$?" in
0) ;;
*) die "You must be root to $action this." ;;
esac
}
# Make sure we have permission and a safe tempfile.
needroot
systype=$(uname -s | tr A-Z a-z)
tmp=$(mktemp -q "/tmp/$tag.XX")
case "$?" in
0) test -f "$tmp" || die "$tmp: tempfile not created" ;;
*) die "$tmp: mktemp failed" ;;
esac
# Real work starts here. Check for OS-specific instructions.
case "$systype" in
freebsd)
( swapoff -a && swapon -a ) >> $tmp 2>&1
;;
linux)
mem=$(free | awk '/Mem:/ {print $4}')
swap=$(free | awk '/Swap:/ {print $3}')
if test "$mem" -lt "$swap"; then
logmsg "not enough RAM to recover swap, nothing done"
else
( swapoff -a && swapon -a ) >> $tmp 2>&1
fi
;;
esac
# Cleanup.
test -s "$tmp" && logfile $tmp
rm $tmp
exit 0
Re: Work environment
On Fri, Nov 24, 2023 at 03:28:34AM -0500, Darac Marjal wrote: > On 23/11/2023 04:34, William Torrez Corea wrote: > > Why the people use two desktops and one laptop? > Without any context, it's hard to answer. But there are some possibilities: > > * Regardless of any other factor, desktops are bigger than laptops, so > there is room for more hard drives, optical drives, more PCI cards etc. It gives me a big monitor, a good mouse, and a keyboard that doesn't suck. -- Karl Vogel I don't speak for anyone but myself Drunk Grizzlies Keep Getting Hit By Trains In Montana --"Cowboy State Daily" headline, 4 Nov 2023
Re: Too much log for sudo.
On Thu, Oct 12, 2023 at 11:22:00AM -0400, Erwan David wrote: > I use a script to run borg backup. For it to be able to backup files that > only root may read, i use sudo --preserv-env=BORG_REPO,BORG_PASSPHRASE. > > However I see that in the logs the VALUE of the env variable is loggued. How > to change this? You can either run "sudo -E" to push the entire environment through without echoing any values in the sudo log, or play some games by re-invoking the script with a clean environment. My environment with the BORG variables: me% export BORG_REPO=/path/to/repo me% export BORG_PASSPHRASE='horse battery' me% env | sort ATTRIBUTION=%f wrote: BLOCKSIZE=1m BORG_PASSPHRASE=horse battery BORG_REPO=/path/to/repo EDITOR=vim [diaper-load of other variables] XDG_CACHE_HOME=/home/vogelke/.cache XDG_CONFIG_HOME=/home/vogelke/.config XDG_DATA_HOME=/home/vogelke/.local/share XDG_RUNTIME_DIR=/home/vogelke/.local/run XDG_STATE_HOME=/home/vogelke/.local/state Script to see if I'm running under a regular environment -- if so, restart the same script with a bare minimum environment plus the BORG variables: me% cat tst #!/bin/bash #
Re: Sunrise and Sunset from terminal
On Fri, Sep 29, 2023 at 12:48:56PM -0400, Bruno Kleinert wrote: > Am Samstag, dem 23.09.2023 um 23:51 +0200 schrieb s...@gmx.com: > > Is there a way to get sunrise and sunset time from command interpreter? > > I want to use its output for a script! > > in case you're looking for a possibility to execute commands at sunset > and/or sunrise, I use remind for this (apt install remind) to trigger my > roller blinds. Pro: Does not rely on any external web service. Remind is an incredibly useful program. Give it your latitude and longitude, and it can tell you sunrise and sunset times. My coordinates: https://www.google.com/search?q=dayton+ohio+latitude+longitude+degrees+minutes+seconds Dayton, OH, USA is located at GPS coordinates of 39d 45m 32.2164s N and 84d 11m 29.7780s W. Version of remind: me% remind -V 2>&1 | grep Copyright REMIND 04.02.04 (English version) Copyright 1992-2023 Dianne Skoll Remind script: me% cat sunrise-sunset.rem ; Show sunrise/sunset SET $LatDeg 39 SET $LatMin 45 SET $LatSec 32 SET $LongDeg 84 SET $LongMin 11 SET $LongSec 29 MSG sunrise at [sunrise(trigdate())], sunset at [sunset(trigdate())] Sunrise/sunset for today: me% remind -h sunrise-sunset.rem '*1' | sed -e '/^$/d' Reminders for Friday, 29th September, 2023 (today): sunrise at 07:30, sunset at 19:22 Sunrise/sunset for the next week: me% remind -h sunrise-sunset.rem '*7' | sed -e '/^$/d' Reminders for Friday, 29th September, 2023 (today): sunrise at 07:30, sunset at 19:22 Reminders for Saturday, 30th September, 2023: sunrise at 07:31, sunset at 19:21 Reminders for Sunday, 1st October, 2023: sunrise at 07:32, sunset at 19:19 Reminders for Monday, 2nd October, 2023: sunrise at 07:33, sunset at 19:18 Reminders for Tuesday, 3rd October, 2023: sunrise at 07:34, sunset at 19:15 Reminders for Wednesday, 4th October, 2023: sunrise at 07:35, sunset at 19:15 Reminders for Thursday, 5th October, 2023: sunrise at 07:36, sunset at 19:12 -- Karl Vogel I don't speak for anyone but myself. German: "Er hat da geparkt, wo ihm die Karre vom Arsch gefallen ist." English: "He parked where the car fell off his ass." --Reddit comment about stupid/illegal parking
Re: [a bit OT] Automate a (G o o g l e) search from a list of strings
On Tue, Sep 19, 2023 at 10:42:14AM -0400, steve wrote:
> I have a list of 200 keywords and would like for every one to launch a
> search on a specific website and put the result(s) in a file [...]
Take a list of words and turn it into a single Google query matching any
of them:
me% cat keywords
corsairs
buccaneers
privateers
Desired query:
https://www.google.com/search?q=corsairs+OR+buccaneers+OR+privateers
Script:
me% cat search
#!/bin/sh
export PATH=/usr/local/bin:/bin:/usr/bin
set -o nounset
umask 022
query=$(tr "[:space:]" "+" < keywords |
sed -e 's/ /+OR+/g' -e 's/+$//' -e 's/+/+OR+/g')
curl -s -L -o pirate.htm "https://www.google.com/search?q=${query}";
ls -l pirate.htm
exit 0
Results:
me% ./search
-rw-r--r-- 1 vogelke 220487 19-Sep-2023 21:33:21 pirate.htm
If you want results in text form, lots of programs can do that:
me% lynx -cookies -accept_all_cookies -dump -width 80 pirate.htm
Tampa Bay Buccaneers: Buccaneers Home
www.buccaneers.com
The official source of the latest Bucs headlines, news, videos, photos,
tickets, rosters, stats, schedule and gameday information.
People also ask
What is the difference between a pirate and a corsair?
What kind of pirate is a corsair?
What were Spanish pirates called?
What is the history of corsairs?
Pirates, Privateers, Corsairs, Buccaneers: What's the Difference?
www.britannica.com > Demystified > Geography & Travel
Corsairs were essentially privateers, although the term corsair carried
an added religious connotation because the conflict was between Muslim
and Christian...
HTH.
--
Karl Vogel I don't speak for the USAF or my company
Leonardo diCaprio in Titanic: Despite acting like a plank of wood,
he still can't float. --"Top Ten Most Welcome Movie Deaths", Stylus Magazine
Re: door bell like sound effect
On Wed, Aug 30, 2023 at 07:55:14AM -0400, songbird wrote: > Karl Vogel wrote: > ... > > If nothing else, it's faster to run "locate" and look for file extensions; > > running "file" on that much crap took nearly 9 hours. > > do you have SSDs or spinning rust? I have a 256-Gb SSD and two mirrored Western Digital Blue 1.8-Tb drives. About 2 million files are on SSD and the rest are on rust. I used "file" v5.45 built from source, which does a nice job but is IO- and CPU-intensive. > when i just did this: > # find / -type f | wc -l > it took all of 24 seconds for the 2.4 million files found. Generating hashes for SSD files is faster than getting the filetype; it takes about 17 minutes for 3.6 million files (153 Gbytes). I like the Blake-2 hash cuz it's fast as hell, among other things: #!/bin/ksh #> $work done # Store hashes for SSD datasets. # The hash file is sorted by filename to make comparisons easier. logmsg "running b2sum" fdbdir=$(date '+/var/fdb/%Y/%m%d') sort -z $work | xargs -0r b2sum -l 128 > "$fdbdir/zroot.sum" rm $work exit 0 Useful for finding changed files -- security, backups, etc. > what script did you use? #!/bin/ksh # $work logmsg finish mv $work filetypes exit 0 -- Karl Vogel I don't speak for the USAF or my company Assisted in daily preparation of large quantities of consumable items in a fast-paced setting. (Translation: Short-order cook) --from a list of resume' blunders
Re: door bell like sound effect
Out of morbid curiosity (and boredom), I started wondering what types of audio files I had on my systems. I ran "file --mime-type" on 6.8 million files, looked for "audio/whatever" and got the file extensions. Extension MIME-type --- .8svx audio/x-aiff .aif audio/x-aiff .aifc audio/x-aiff .aiff audio/x-aiff .ape audio/x-ape .arm audio/amr .auaudio/basic .flac audio/flac .m4a audio/x-m4a .mp3 audio/mpeg .mpc audio/x-musepack .oga audio/ogg .ogg audio/ogg .opus audio/ogg .raaudio/x-pn-realaudio .voc audio/x-unknown .wav audio/x-wav If nothing else, it's faster to run "locate" and look for file extensions; running "file" on that much crap took nearly 9 hours. -- Karl Vogel I don't speak for anyone but myself My doctor told me to stop having intimate dinners for four, unless there are three other people.--Orson Welles
Re: door bell like sound effect
On Tue, Aug 29, 2023 at 10:53:39PM -0400, gene heskett wrote: > And that is the problem, and why I read thru all those man ages trying to > find a way to make it log what it did. Sadly no. Install and configure file auditing on Debian: https://www.daemon.be/maarten/auditd.html Auditing can help you find anything trying to write to your sound device. Look at these manpages: auditd.conf (5) audit.rules (7) audispd (8) ausearch (8) aureport (8) auditctl (8) augenrules (8) To find your sound cards and/or devices: https://wiki.debian.org/Sound https://wiki.debian.org/SoundCard https://wiki.debian.org/PulseAudio I don't have a Debian system to play with, but in the (good|bad) old days, we had a /devices directory with all sorts of weirdness inside. If you have one of those, try find /devices -print | grep sound That might point you to an actual device name. -- Karl Vogel I don't speak for anyone but myself. Comment: One of my friends drank half a bottle of rum and refilled it with a bodily fluid of similar color. Reply: Your friend should see a doctor and drink more water. --seen on Reddit, 27 Aug 2023
Re: Using the bash shell: determine if the root user used 'sudo -i'
On Sat, Aug 26, 2023 at 12:09:57PM -0400, Tom Browder wrote:
> Excellent mind-reading, Greg! So to use your line I will put in that dir:
> "cd /required-dir || exit"
>
> Thanks so much. And thanks to all others who responded.
If you're running bash, the safest way to find your current working
directory is capturing the output from /bin/pwd. Symlinked directories
can surprise you:
me$ cd
me$ ls -ldF today
lrwxr-xr-x 1 me mis 18 Aug 26 00:03 today@ -> notebook/2023/0826
me$ cd today
me$ pwd
/home/me/today
me$ /bin/pwd
/home/me/notebook/2023/0826
me$ echo $PWD
/home/me/today
If you want to know why you had an early exit:
me$ cat try
#!/usr/bin/env bash
# try: test logging.
export PATH=/usr/local/bin:/bin:/usr/bin
set -o nounset # check for unbound variables.
tag=${0##*/}
umask 022
# Test file descriptor 2 for interactive or cron use.
test -t 2
case "$?" in
0) logmsg () { echo "$(date '+%F %T') $tag: $@"; } ;;
*) logmsg () { logger -t $tag "$@"; } ;;
esac
warn () { logmsg "WARN: $@" ; }
die () { logmsg "FATAL: $@"; exit 1 ; }
# Real work starts here.
case "$#" in
0) die "need a directory" ;;
*) dir="$1" ;;
esac
test -d "$dir" || die "$dir: not a directory"
cd "$dir" || die "$dir: cannot cd"
cwd=$(/bin/pwd)
logmsg "start working in $cwd"
exit 0
On FreeBSD, you can use "daemon" to run something detached from the
controlling terminal, which simulates running a cron job:
me$ ls -ldF /etc /var/authpf
drwxr-xr-x 27 root wheel 120 26-Aug-2023 07:55:02 /etc/
drwxrwx--- 2 root authpf 2 05-Jul-2019 00:45:45 /var/authpf/
me$ ./try /etc
2023-08-26 18:31:54 try: start working in /etc
me$ daemon -f $PWD/try /etc
me$ daemon -f $PWD/try /var/authpf
me$ tail -2 /var/log/syslog
Aug 26 18:19:17 myhost try: start working in /etc
Aug 26 18:19:19 myhost try: FATAL: /var/authpf: cannot cd
Hope this helps.
--
Karl Vogel I don't speak for anyone but myself.
Oh, my darlin' had bronchitis and she barfed up half a lung,
what came up looked quite amazing when she rolled it on her tongue.
--sung to the tune of "My Darling Clementine"
Re: git setup
On Mon, Aug 21, 2023 at 11:01:21PM -0400, Russell L. Harris wrote:
> I write articles for publication. I typically spend anywhere from
> several hours to many days on each article. It is frustrating to work
> for an hour or two on a paragraph or a page and then accidentally to
> erase what I have written.
Heard that.
> The second repository is my backup; it resides on another machine.
> Several times a day, I SSH into the backup machine and pull the
> working repository. It would be nice to be able to push from WORKING
> to BACKUP, eliminating the need to SSH.
I'd set up a post-commit hook on the production system. Have it SSH to
your backup system and do a pull automatically whenever you commit a change.
I made an example project to try it:
me% mkdir example
me% cd example
me% git init
Test file:
me% date > testing
me% git add testing
me% git commit -m 'Testing a new file"
me% git log
d5fe2ce 2023-08-22 00:09:21 -0400 vogelke - Testing a new file
Here's a test script to connect from my backup system ("bkup") and simulate
pulling my working repo. I have an SSH public/private key set up for
passwordless login:
me% cat try
#!/bin/sh
export PATH=/usr/local/bin:/bin:/usr/bin
ssh -q -c aes128-...@openssh.com -i $HOME/.ssh/bkup_ed25519 \
bkup "logger -t autopull git pull whatever"
exit 0
Here's the remote system log when I run this from the command line:
me@bkup% tail -1 /var/log/messages
Aug 22 00:24:44 bkup autopull[60592]: git pull whatever
I copied the test script ("try") to .git/hooks/post-commit and ran a commit:
me% git add try
me% date; git commit -m 'Added hook'
Tue Aug 22 00:31:27 EDT 2023
[master 42cb917] Added hook
1 file changed, 6 insertions(+)
create mode 100755 try
me% git log
42cb917 2023-08-22 00:31:27 -0400 vogelke - Added hook
d5fe2ce 2023-08-22 00:09:21 -0400 vogelke - Testing a new file
Here's the system log on the remote system showing my hook ran when the
commit was done:
me@bkup% tail -1 /var/log/messages
Aug 22 00:31:27 bkup autopull[80162]: git pull whatever
Replace the "logger ..." part of post-commit to run "git pull" or whatever.
Hope this helps.
--
Karl Vogel I don't speak for the USAF or my company
There's no kill like overkill.
--Reddit post about using .45 ACP rounds to dispose of old disk drives
Re: xterm font and other options
On Sun, Aug 20, 2023 at 10:38:34PM -0400, Max Nikulin wrote:
> On 20/08/2023 14:55, Karl Vogel wrote:
> > #!/bin/sh
> ...
> > # -fa 'xft:...' font size and weight
> ...
> > ( $XTERM $geo $topts -fa "$FONT" -title "Remote" ) &
>
> Xterm configuration options may be put to ~/.Xresources, e.g.
>
> xterm*VT100.faceName: ...
>
> I am curious if there are actual advantages of usage a wrapper script
> instead of xresources.
For me, being able to select or change a font based on an environment
variable is very convenient.
The script I included is simplified because I didn't want the post to
get too long. My production version has other conveniences:
# Don't override COLUMNS and LINES if already set; when my eyes are
# tired, I use an xterm with characters two pixels larger:
## FONT=xft:Cascadia:pixelsize=22:bold LINES=35 xt
: ${COLUMNS=80}
: ${LINES=40}
I can check a font and set LINES, COLUMNS, or geometry on the fly without
having to mess with any configuration options.
--
Karl Vogel I don't speak for the USAF or my company
Tent poles are not for pole dancing. Please find
alternative ways to disappoint your father. --seen on boredpanda.com
Re: Looking for a good "default" font (small 'L' vs. capital 'i' problem)
On Sat, Aug 19, 2023 at 03:29:22PM -0400, Christoph K. wrote:
>
> I'm unsatisfied with the default sans font in debian for use in the
> graphical user interface (in my case XFCE).
I use BSD and Linux, and my eyesight sucks. For console work (23" monitor
that's about 2 feet away) I use an Xterm with one of the following fonts
(in order of preference):
* xft:Menlo-Regular:pixelsize=20:bold
* xft:Cascadia:pixelsize=22:bold
* xft:Bitstream Vera Sans Mono:pixelsize=21:bold
For browsing (Firefox), my "prefs.js" file holds:
user_pref("browser.display.use_document_fonts", 0);
user_pref("font.default.x-western", "sans-serif");
user_pref("font.internaluseonly.changed", false);
user_pref("font.minimum-size.x-western", 18);
user_pref("font.name.monospace.x-western", "DejaVu Sans");
user_pref("font.name.sans-serif.x-western", "sans-serif");
user_pref("font.name.serif.x-western", "DejaVu Serif");
user_pref("font.size.fixed.x-western", 18);
user_pref("font.size.variable.x-western", 18);
Others I've liked:
* xft:Edlo:pixelsize=21:bold
* xft:FiraMono-Regular:pixelsize=22:bold
* xft:Inconsolata-Bold:pixelsize=25:bold
* xft:Meslo LG S:pixelsize=20:bold
* xft:Meslo LG S:pixelsize=21:bold
* xft:UbuntuMono-B:pixelsize=25:bold
If you get your FONT setting from the environment, it's easy to experiment:
me% echo $FONT
xft:Cascadia:pixelsize=20:bold
This script starts a new xterm with some tweaks to make it a little nicer:
#!/bin/sh
#https://invisible-island.net/archives/xterm/xterm-384.tgz
https://invisible-island.net/archives/xterm/xterm-384.tgz.asc
To build, unpack the source:
dest=/usr/local
export TERMINFO=/usr/local/share/terminfo
./configure \
--disable-setgid \
--disable-setuid \
--enable-256-color \
--enable-narrowproto \
--mandir=$dest/man \
--with-x \
--with-own-terminfo=$TERMINFO
make
make check
make install
It comes with a nice terminfo file. I've had problems with "tic" for
ncurses >= version 6, so I use the ncurses-5.9 version to compile it:
root# tic59 -V
ncurses 5.9.20110404
root# tic59 -s -o $TERMINFO terminfo
Hope this gives you some ideas.
--
Karl Vogel / vogelke AT pobox DOT com / I don't speak for anyone but myself
The Beatles: "I Get By with a Little Help From Depends"
--Re-released hits for an aging audience
Re: Recommendations for a UPS?
>> On Tue, 01 Aug 2023 03:47:28 -0400, >> David may have said: > What everybody seems to be doing is catering to surge, when a low > spike can do just as much damage. Both need to be protected against, > so any protective appliance selection has to consider that. Liebert protects against power surges, brownouts, blackouts or drops in voltage. That's why they're a little more expensive. -- Karl Vogel / vogelke AT pobox DOT com / I don't speak for anyone at the moment Skills-wise they suck harder than a black hole with daddy issues. --Review of Accenture Senior Management
Re: Recommendations for a UPS?
>> On Mon, 31 Jul 2023 13:24:36 -0400, >> Tom Browder may have said: > All the reviews I've seen on Amazon for smaller capacity UPSs > for APC and Tripp Lite are not that great (I usually concentrate > on the one- and two-star reviews). > Any recommenndations from fellow Debian folks? > Thanks. I give a solid vote to Liebert. I had a near-miss lightning strike a few nights ago, and all it did was make my display go out for about a second. It came right back, session intact, didn't lose a thing. -- Karl Vogel / vogelke AT pobox DOT com / I don't speak for anyone at the moment Why Trick or Treating is Better than Sex #9: If you get tired, you can wait 10 minutes and go at it again.
Re: Expect script does not work on crontab
>> On Tue, 13 Mar 2012 13:59:35 +0530, >> Bilal mk said: B> The following script is working fine executing from shell. But does not B> work running on crontab. How to fix this crontab issue? B> #!/usr/bin/expect B> spawn /usr/bin/rsync -avz --delete /var/app/ user@somedomain:/var/app/ B> expect "Password:" B> send "somepassword\r" B> interact I believe the "interact" line is messing you up; when "expect" sees that, it wants a terminal or PTY to play with, and cron doesn't give it one. -- Karl Vogel I don't speak for the USAF or my company I smile because I don't know what's going on.--bumper sticker -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120313184717.e6d36b...@kev.msw.wpafb.af.mil
Re: Cron, > 1 month but < 2 months
>> On Sun, 11 Mar 2012 01:42:25 -0600, Christofer C. Bell wrote: C> According to this page[1], the following works for that: C>* * */45 * * your_scheduled_task C> I've not tried it myself. I'd use a timestamp file in conjunction with find; what happens when your 45-day interval crosses Jan 1st? For example: me% date Sun Mar 11 21:51:21 EDT 2012 me% touch -d '45 days ago' timestamp me% ls -l --time-style='+%d-%b-%Y %T' timestamp -rw-r--r-- 1 me me 0 26-Jan-2012 20:51:27 timestamp me% find timestamp -daystart -mtime +1 -print timestamp me% find timestamp -daystart -mtime +44 -print timestamp me% find timestamp -daystart -mtime +45 -print # no output Run something like this every day: #!/bin/bash ts=/time/stamp/file interval=45 set X $(find $ts -daystart -mtime +$interval -print) shift case "$#" in 0) logger 'starting' ;; *) logger 'skipping'; exit 0 ;; esac # run your php script or whatever ... touch $ts Another advantage: you can decide whether to "touch $ts" based on the return code from whatever you're trying to run. If it fails, do you want to alert someone and run it tomorrow or wait another 45 days? Reschedule by setting the $ts modtime instead of dorking around with cron. -- Karl Vogel I don't speak for the USAF or my company Bumper-sticker on Mel Gibson's car: "Swerve If You Love Jesus" -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120312021336.658d0b...@kev.msw.wpafb.af.mil
Re: How to keep track of files installed from sources?
>> On Sat, 18 Feb 2012 21:53:49 +0100, >> Csanyi Pal said: C> I want to install GNUstep from sources on my Debian GNU/Linux SID sytem. C> How can I know where come installed files after I did run: ./configure C> && make && sudo -E make install commands? C> this is important if I decide later to uninstall files installed from C> sources. If you want a list of what files were installed, the easiest way is probably to do the configure/make/make install dance twice. The first time, install under a disposable directory: root# mkdir /tmp/local root# cd /your/source/directory root# ./configure --prefix=/tmp/local --whatever root# make && make install If you're planning on installing under /usr/local: root# find /tmp/local -printf "%y %p %m %u %g %l\n" | sort -k2,2 | sed -e 's!/tmp/local!/usr/local!' > list-of-installed-files root# rm -rf /tmp/local root# make distclean # or make realclean, start from scratch root# ./configure ... # use real options this time root# make && make check && make install "find" gives the filetype, path, mode, owner, group, and an optional 6th field if the file is a symbolic link; "sort" gives results sorted by filename. Some sample output: d /usr/local/bin 755 root bin f /usr/local/bin/antiword 755 root bin f /usr/local/bin/authlog 755 root bin f /usr/local/bin/avg 755 root bin l /usr/local/bin/bootlog 777 root bin authlog d /usr/local/lib 755 root bin ... Here, /usr/local/bin/bootlog is a symlink to /usr/local/bin/authlog. If you're paranoid, either save the output from "make install" or do a separate check to make sure nothing was written outside of /usr/local: root# find / /usr -xdev -mtime -1 -ls -- Karl Vogel I don't speak for the USAF or my company I put instant coffee in a microwave and went back in time. --Steven Wright -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120309192354.29d58b...@kev.msw.wpafb.af.mil
Re: Understanding the -depth option of find?
>> On Thu, 8 Mar 2012 17:06:54 + (UTC), said: C> I make the backup for my /home directory with plain "tar". What are the C> benefits/drawbacks of using "cpio" instead? Older versions of tar wouldn't read from stdin or any other list of files, and they also had really short limits on filename length. Some still do. I like cpio because it accepts output from "find", a list of files, or any intermediate filter. I prefer pax because in addition, it'll accept cpio, ustar, or its own internal format. Other advantages include consistent behavior and command-line options on FreeBSD, Linux, and Solaris. Doing things like hourly incremental/differential backups is a snap: root# touch /backup/etc/drive1 [ ...an hour later... ] root# cd /drive1 root# dest="$(date '+/backup/drive1/%Y/%m%d/%H%M')" root# mkdir -p $dest root# find ./userstuff -depth -newer /backup/etc/drive1 -print | pax -rwd -pe $dest In this case, the "-depth" option ensures correct timestamps on all created directories, and "-pe" preserves owner/group/mode/modtime stuff. -- Karl Vogel I don't speak for the USAF or my company Heck, if an omniscient psycho ex offered me discounted, in-stock products I was going to buy anyway, I might just keep them around. That's everything Amazon tries to be already, anyway. --slashdot response to Target data-mining for pregnant customers -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120308215615.05c7bb...@kev.msw.wpafb.af.mil
Re: history configuration
>> On Wed, 29 Feb 2012 00:19:19 +0800, lina said:
L> How to configure to make each terminal share one "history"? I found
L> each terminal saves its own history and no sharing. Is it possible to
L> let them share?
You can if you don't mind stretching the definition of "share". Have a
look at this page to see how to save your command history to syslog:
http://jablonskis.org/2011/howto-log-bash-history-to-syslog/
Here's an example. I added this to the end of ~/.bashrc to save history
to my own file and the system log:
export PROMPT_COMMAND='history -a >(tee -a ~/.bash_history |
logger -p local2.notice -t "$USER [$$] $SSH_CONNECTION")'
This sends my history to facility "local2" with priority "notice".
I fixed syslog so local[25] messages only show up in one place; I already
use local5 for output from a backup program. Here's the relevant part
of /etc/syslog.conf -- backslashes must be at the end of the lines:
# Log anything (except mail) of level info or higher.
# Don't log private authentication messages!
*.info;kern.none;mail.none;\
authpriv.none;auth.none;local2.none;\
local5.none;cron.none /var/log/messages
# Auth logfiles are restricted.
authpriv.* /var/log/secure
auth.* /var/log/authlog
# Local logs.
local0.* /var/log/local0log
local1.* /var/log/local1log
local2.* /var/log/local2log
local3.* /var/log/local3log
local4.* /var/log/local4log
local5.* /var/log/local5log
local6.* /var/log/local6log
# Cron stuff
cron.* /var/log/cron
Here's what my history looks like:
me% tail -3 /var/log/local2log
Mar 7 22:09:22 svr1 vogelke [149] 1.2.3.4 46555 5.6.7.8 22: which logger
Mar 7 22:09:31 svr1 vogelke [149] 1.2.3.4 46555 5.6.7.8 22: echo $PATH
Mar 7 22:09:44 svr1 vogelke [149] 1.2.3.4 46555 5.6.7.8 22: dir
/var/log/local*
me% tail -3 $HOME/.bash_history
which logger
echo $PATH
dir /var/log/local*
"1.2.3.4" is where I'm connecting from, "5.6.7.8" (aka "svr1") is the
host I'm connecting to via SSH. If you use the Z-shell instead of bash,
the method is pretty similar -- add this function to the bottom of your
~/.zshrc file:
# Log commands, return codes and the current working directory.
# Based on:
#
# http://blogs.sun.com/chrisg/entry/logging_commands_in_korn_shell
# Logging commands in korn shell
# Chris Gerhard
# Thu, 2 Mar 2006 09:47:29 -0500
#
# Don't try to put local on the x= line. If you do,
# any command arguments (i.e., "ls -la") will throw an error:
# precmd:local:2: not an identifier: -la
precmd () {
typeset -i stat=$?
local x
x=$(fc -ln -1)
local d="$(/bin/pwd)"
logger -p local2.notice -t "$LOGNAME $$" $stat: $d: \($x\)
}
In order to really share this, you'd probably have to write something to
grub around in the syslog file.
--
Karl Vogel I don't speak for the USAF or my company
She acts like she's on the phone to avoid certain customers. One time it rang.
--Jimmy Fallon, #mycoworkeriscrazy tweets, 2 Mar 2012
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120308040126.7ec1cb...@kev.msw.wpafb.af.mil
Re: [OT] A quick Q: how to check the webpage building time
>> On Mon, 5 Mar 2012 23:28:12 +0800, >> lina said: L> Sometimes I wish to know the info. on some webpage, which is latest news L> or 7 years ago. How can I know the building time of those information. For static files, either "curl" or "HEAD" can do this. For example: % curl -I http://www.hcst.net/~vogelke/pictures/2012/0105-new-year-wishes/p04.jpg HTTP/1.1 200 OK Date: Tue, 06 Mar 2012 03:49:27 GMT Server: Apache Last-Modified: Thu, 05 Jan 2012 20:04:57 GMT ETag: "414a07-3a93-4b5cd7246ac40" Accept-Ranges: bytes Content-Type: image/jpeg Content-Length: 14995 A dynamically-generated page will return the current time. Also, the server may be configured to NOT return the modtime for (say) an index.html file, even if it's static. -- Karl Vogel I don't speak for the USAF or my company For a nosebleed: Put the nose much lower then the body until the heart stops. --submitted to science and health teachers, Jr High thru college -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120306035551.74d65b...@kev.msw.wpafb.af.mil
Re: Question about ssh passwords and backup software
>> Alex Mestiashvili wrote: A> I would simply use a passwordless ssh-key with a wrapper on the remote A> side which allows to run only the backup command . >> On Mon, 13 Feb 2012 18:51:26 +, >> Chris Davies said: C> I'd agree with this, but use passwordless public/private keys with a C> restricted target command: C> command="backup-service",no-pty,no-port-forwarding ssh-rsa BLAHBLAH... If the box from which you're copying has a static hostname or IP address, include that as well: from="1.2.3.4",command="backup-service",no-pty,no-port-forwarding ... -- Karl Vogel I don't speak for the USAF or my company Why Trick or Treating is Better than Sex #10: You are guaranteed to get at least a little something in the sack. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120214013943.97494b...@kev.msw.wpafb.af.mil
Re: [OT] how to take care of hand
R> On Mon, 6 Feb 2012 14:47:26 -0500 (EST) (Karl Vogel) wrote: K> I correct shoulder pain by doing 5-10 very slow pushups just before bed, K> or by holding a dumbbell over my shoulder and letting my hand dangle K> over the back of the couch while watching TV. >> On Mon, 6 Feb 2012 20:22:29 +, >> richard said: R> you are referring to 10 push ups with each arm aren't you? Nope. By "slow", I mean 5 seconds down, hold for a second, then 5 seconds up for one pushup; if you need more, knock yourself out. It's not intended to replace your morning jog with Seal Team Six. ...I'm such a snot, it's a wonder I have any friends at all... R> but I'm in the state of the beggar in "life of Brian". Just remember, we're *all* individuals. -- Karl Vogel I don't speak for the USAF or my company This punishment is not boring and pointless. --written on blackboard by Bart Simpson -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120208193510.ce067b...@kev.msw.wpafb.af.mil
Re: [OT] how to take care of hand
>> On Fri, Feb 3, 2012 at 18:57, lina wrote: L> are there some suggestions about how to take care of hand? I've had similar problems with my wrists and shoulders, and the one thing that helped most was taking 3 minutes a day to do a set of complete range-of-motion wrist exercises with really small dumb-bells (5 pounds or less). Holding the weights palm-up and rolling them from fingertip to palm and back really did the trick. I correct shoulder pain by doing 5-10 very slow pushups just before bed, or by holding a dumbbell over my shoulder and letting my hand dangle over the back of the couch while watching TV. A few minutes of that and my arm turns to spaghetti, but it feels better in the morning. I've been told that not getting enough B-vitamins drops the effectiveness of other supplements you take; they'll be washed out before the body can use them, and they tend to get eaten up faster when under stress. I also take cod-liver oil to keep my joints nice and gushy. Your doctor may disagree. -- Karl Vogel I don't speak for the USAF or my company ANTI-TALKSIDENT: A spray carried in a purse or wallet to be used on anyone too eager to share their life stories with total strangers in elevators. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120206194726.b731db...@kev.msw.wpafb.af.mil
Re: how to check the server can be connected or not
>> On Fri, 3 Feb 2012 11:49:22 +0800, lina said: L> Hi, a server is closed, but will start in future, sooner or later, I L> wish to access in the first time it's open. I tried: L> while ( ssh server ) do sleep 60 done L> wall < server_open.txt Instead of polling the server, could you have it send a mail message to a special address when starting SSH? If you're running the while-loop on your workstation, set up an extension address like lina-w...@your.workstation.com which forwards incoming mail from the SSH server to /usr/bin/wall. If you use Postfix, see the http://www.postfix.org/local.8.html sections on ADDRESS EXTENSION and EXTERNAL COMMAND DELIVERY. -- Karl Vogel I don't speak for the USAF or my company The patient is tearful and crying constantly. She also appears to be depressed.--actual comment on hospital chart -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120206194121.43109b...@kev.msw.wpafb.af.mil
Re: how to check 400 files exist
>> On Thu, 19 Jan 2012 00:34:50 +0800, >> lina said: L> I am choked by how to check more than 400 files exist or not, if not, L> sleep, but once it's all generated, continue do something. Those files L> have a common feature: sys_em_$i.txt The script below doesn't do any glob-expansion of filenames, so it won't barf if you have a lot of files. It assumes you're in the same directory as the files you're checking. -- Karl Vogel I don't speak for the USAF or my company A raccoon tangled with a 23,000 volt line today. The results blacked out 1400 homes and, of course, one raccoon.--Steel City News -- #!/bin/bash export PATH=/usr/local/bin:/bin:/usr/bin case "$#" in 0) echo need a number; exit 1 ;; *) my_number=$1 ;; esac while : do if [ $(find . -name 'sys_em_*.txt' -print | wc -l) == "$my_number" ] then echo 'success' break fi echo 'test' sleep 5 done echo 'done' exit 0 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120120200208.79090b...@kev.msw.wpafb.af.mil
Re: How to do this, fold + split
>> On Fri, 30 Dec 2011 15:09:13 + (UTC), >> T o n g said: T> I want to split a file every ### of chars. Is it possible not to split T> on the word but word boundaries? Yup, for certain meanings of "split". GNU fmt is part of "coreutils". Here's sample.txt (rulers added for readability): me% cat sample.txt *1*2*3*4*5*6*7* Here is what I normally do. Here is a skeleton script that creates a temporary file, redirects all output there, then if there is output it emails it off. There are many ways to do this and I tinker it as needed. me% gfmt -65 sample.txt *1*2*3*4*5*6*7* Here is what I normally do. Here is a skeleton script that creates a temporary file, redirects all output there, then if there is output it emails it off. There are many ways to do this and I tinker it as needed. me% gfmt -75 sample.txt *1*2*3*4*5*6*7* Here is what I normally do. Here is a skeleton script that creates a temporary file, redirects all output there, then if there is output it emails it off. There are many ways to do this and I tinker it as needed. If you want one word per line, which can be very handy: me% gfmt -1 sample.txt Here is what [...] as needed. -- Karl Vogel I don't speak for the USAF or my company A map with a bullet hole in it is still a map. A computer with a bullet hole in it is a paper-weight. --why soldiers tend to like paper -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120106201008.13767b...@kev.msw.wpafb.af.mil
Re: fastcgi too slow in debian squeeze
>> On Tue, 20 Dec 2011 12:26:44 +0530, >> "J. Bakshi" said: J> I have been suffering with the slow performance of my apache server J> (2.2.16) running on debian wheezy. I have finally pointed out it is the J> fastcgi which makes the web server so slow. But I really need it as I am J> also hosting some sites which do require php 5.2. The symptoms of having J> the fastcgi is very annoying. J> 1 All sites need 30 or so sec. to open and after that the browsing is J> again normal for that site. Whenever I've had a 30-sec-delay problem opening a site, it's almost always been something DNS-related. Do you have any other apps requiring external host resolution that are seeing similar delays? A caching DNS resolver usually fixes things, and it's not hard to set up. -- Karl Vogel I don't speak for the USAF or my company You have experienced an Action Request System Error (ARSE). --Remedy error message, much funnier if you're in England -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111221025801.23a8ab...@kev.msw.wpafb.af.mil
Re: Crazy (?) idea: screen locker with simpler password
>> On Mon, 7 Nov 2011 23:40:24 -0700, >> Bob Proulx said: B> I would create a script that edited the /etc/shadow file directly and B> manipulated the encrypted passwords. Then the clear text would never B> need to exist in any form. Only the encrypted form of the password is B> needed. Use a script to swap between two different encrypted forms. If you don't add or delete a lot of accounts or modify /etc/passwd frequently, could you change your password to the high-security one, copy /etc/shadow to /etc/shadow.hi, then change it back and copy /etc/shadow to /etc/shadow.lo? Then your password-changer could be: alias hisec='sudo cp -p /etc/shadow.hi /etc/shadow' alias losec='sudo cp -p /etc/shadow.lo /etc/shadow' -- Karl Vogel I don't speak for the USAF or my company It's impossible to sneeze with your eyes open. --item for a lull in conversation -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2009014634.bd5eab...@kev.msw.wpafb.af.mil
Re: linux friendly hardware inquiry/advise
>> On Wed, 19 Oct 2011 22:24:37 +1000, yudi v said: Y> Building a new PC, will be used for coding mostly, running 2-3 VMs Y> simultaneously. Debian Linux will be the host OS. I need some Y> suggestions on Motherboards. Which manufacturer has good linux support? The Ars Technica Guides are great if you're building your own box. The system guides go through parts for a budget box, a hot-rod (better than average) and a God box (top-notch, hopefully with someone else's money). The most recent system guide I've seen was last March. http://arstechnica.com/guides/ http://arstechnica.com/gadgets/guides/2011/03/ars-system-guide-march-2011-edition.ars -- Karl Vogel I don't speak for the USAF or my company I've seen people with new children before, they go from ultra happy to looking like something out of a zombie film in about a week. --Alan Cox about Linus Torvalds after his 2nd daughter -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111019210241.bc4a9b...@kev.msw.wpafb.af.mil
Re: Formatting external HDD
>> On Tue, 18 Oct 2011 17:54:32 +0100, Lisi said: L> I have just acquired a one T HDD for use as an external HDD. I now need L> to decide how to partition it. [...] Would it be feasible to have one L> large partition on the drive, and then use directories rather than L> partitions for the different back-ups? Yes. I've been using 1.5Tb Seagate drives for a backup server, and they work fine with one large or two smaller partitions: Filesystem 1M-blocks Inodes Mounted /dev/sda21372701 362774528 /data1 /dev/sdb6 699594 11216896 /data2 /dev/sdb7 699601 11216896 /data3 I used "mkfs.ext3 -J size=400 -i 65536 -m 2" when creating the sdb filesystems, which gave me an extra 27Gb by creating one inode per 64K and only reserving 2% for overflow. I get better performance by using the deadline scheduler and setting vm.swappiness = 10. L> Could I do this with cp (obviously), dd, rsync, Clonezilla, or even L> something I don't know about yet? Sure, cp for the initial copy and then rsync for changes. L> And what filing system? [...] My box is ill, possibly unto death. In that case, now's not the time to experiment. Use something familiar, and play around *after* your stuff is safely backed up. -- Karl Vogel I don't speak for the USAF or my company Salesmen welcome. Dog food is expensive.--seen on a fence -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111019212911.1c538b...@kev.msw.wpafb.af.mil
Re: a quick Q: how to start run a job after another job finished
>> On Thu, 6 Oct 2011 23:12:46 +0800,
>> lina said:
L> I want to run a job after checking another job finished or not, such as
L> another job PID is 5543. If it's finished then run newjob.sh
If you own the other process (or you're root) you can use signal 0 to
verify that it's still alive:
me% sleep 60 &
[1] 10175
me% kill -0 10175 [$? = 0]
me% kill -0 10176
kill: kill 10176 failed: no such process [$? = 1 or 2]
Try this:
pid=5543 # or $1 or whatever
while kill -0 $pid 2> /dev/null; do
sleep 60
done
/path/to/newjob.sh
--
Karl Vogel I don't speak for the USAF or my company
Taxation WITH representation isn't so hot, either.--bumper sticker
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20111007033031.57f74b...@kev.msw.wpafb.af.mil
Re: How do you save passwords for Git/GitHub?
>> On Thu, 21 Jul 2011 12:06:11 -0500,
>> Jason Hsu said:
J> I have a script that uses "git clone" multiple times to download all of
J> the repositories I need for my project Swift Linux. However, I'm asked
J> for my password EVERY TIME the script tries to download a repository. Is
J> there a way to save my password (temporarily) so that I don't have to
J> enter it 20 times?
Do you have TCL and expect installed? If so, expect and autoexpect can
be used to automate just about any interactive process. Here's an example
using a long, random key for your SSH passphrase:
you% cat makekey
#!/bin/ksh
# /dev/null | md5sum`
phrase="$2"
sleep 1
set X `dd if=/dev/urandom count=1 2> /dev/null | md5sum`
phrase="${2}${phrase}"
echo $phrase
# Use it to generate SSH keys.
ssh-keygen -t dsa -b 1024 -f $dir/${name}_dsa -N "$phrase"
ssh-keygen -t rsa -b 1024 -f $dir/${name}_rsa -N "$phrase"
exit 0
you% makekey github
27262a68bd1f633d1702c599f4cd78ad93be2fa7a40a554c15af86a929339df9
Generating public/private dsa key pair.
Your identification has been saved in /home/jhsu/.ssh/github_dsa.
Your public key has been saved in /home/jhsu/.ssh/github_dsa.pub.
The key fingerprint is:
a5:a2:21:82:29:3a:08:d1:bb:e4:38:57:4a:88:cb:bc j...@your.host
Generating public/private rsa key pair.
Your identification has been saved in /home/jhsu/.ssh/github_rsa.
Your public key has been saved in /home/jhsu/.ssh/github_rsa.pub.
The key fingerprint is:
ec:ee:45:bd:73:91:47:24:e4:c6:df:a4:17:b6:ea:fd j...@your.host
you% ls -lF /home/jhsu/.ssh/github*
-rw--- 1 jhsu jhsu 736 Jul 21 15:17 /home/jhsu/.ssh/github_dsa
-rw-r--r-- 1 jhsu jhsu 621 Jul 21 15:17 /home/jhsu/.ssh/github_dsa.pub
-rw--- 1 jhsu jhsu 951 Jul 21 15:17 /home/jhsu/.ssh/github_rsa
-rw-r--r-- 1 jhsu jhsu 241 Jul 21 15:17 /home/jhsu/.ssh/github_rsa.pub
You've got a nice, long key with plenty of entropy. Here's an expect
script that logs you into a host using that key, after you put the public
keys on that host:
1 #!/usr/local/bin/expect -f
2 # SSH connect to some host with userid "jhsu".
3
4 set timeout -1
5 set stty_init -echo
6 spawn ssh -i /home/jhsu/.ssh/github_dsa -c arcfour example.github.org
7 match_max 10
8 expect -exact "Enter passphrase for key '/home/jhsu/.ssh/github_dsa':"
9 send -- "27262a68bd1f633d1702c59...\r"
10 expect "\r"
11 stty echo
12 interact
"Expect" sends strings on your behalf and takes different actions depending
on the responses. Running this script will handle starting the connection
and sending your key; if the login is successful, you should get a shell
prompt, or whatever you're used to seeing when you connect.
line 4: disable any automatic timeouts
line 5: disable echoing so the passphrase doesn't show up
line 6: start the SSH connection to the remote host
line 7: don't try to match more than this many characters
line 8: look for an exact host response of "Enter passphrase..."
line 9: then send the key followed by a return
line 10: look for a return (or some string indicating successful login)
line 11: enable echoing
line 12: let expect know that you'll handle any further interaction
Protect this expect script the same way you do your private SSH keys,
for exactly the same reason.
If the interaction here doesn't match what you normally see, use the
autoexpect program to watch over your shoulder while you do a session by
hand, and it will write a script for you.
--
Karl Vogel I don't speak for the USAF or my company
The last thing I want to do is hurt you, but it's still on my list.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110721205739.b68edb...@kev.msw.wpafb.af.mil
Re: (hopefully perl) API to /etc/network/interfaces?
>> On Tue, 31 May 2011 23:23:19 -0500, >> Mike Mestnik said: M> I'm quite certain this is a some-what usable solution. However it M> doesn't appear to have any means to preserve order. Have a look at the Config::Std module by Damian Conway. It handles a similar problem (reading/writing dot-INI files) but preserves the section order. -- Karl Vogel I don't speak for the USAF or my company Things that never happen in "Star Trek" #7: The Enterprise successfully ferries an alien VIP from one place to another without serious incident. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110602003719.a0b51b...@kev.msw.wpafb.af.mil
Re: Problem with the global weigh of several files
>> On Thu, 26 May 2011 17:12:25 +0200,
>> "Gorka" said:
G> I am looking for the best way of appending commands to obtain the global
G> amount of Mb of the modified files of my PC. Something like this:
G> find ./ -mtime -7 | du -Sch
#!/bin/ksh
# filesize summary; optional argument is starting directory.
export PATH=/usr/local/bin:/bin:/usr/bin
case "$1" in
"") dir="$HOME" ;;
*) dir="$1" ;;
esac
test -d "$dir" || { echo "$dir: not a directory"; exit 1; }
cd $dir|| { echo "$dir: cannot cd"; exit 2; }
dir=$(pwd)
find . -type f -mtime -7 -printf "%s\n" | gawk -v "d=$dir" \
'BEGIN { size = 0; }
{ size += $1; }
END {
printf "%s: %d new files using %.1f Mb\n", d, NR, size/1048576;
}'
exit 0
--
Karl Vogel I don't speak for the USAF or my company
Men are like a fine wine. They start out as grapes and it's up to women
to stomp the sh*t out of them until they turn into something acceptable
to have with dinner.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110526215044.41031b...@kev.msw.wpafb.af.mil
Re: first mkdir takes a long time (on ext3)
>> On 19/05/11 17:01, Hartmut Niemann wrote:
H> It often takes very long time (20s) to mkdir on an ext3 drive
>> Am Fr Mai 20 2011 schrieb Karl Vogel:
K> What does "strace mkdir /some/directory" show?
>> On Fri, 20 May 2011 08:38:41 +0200, said:
H> $ strace mkdir two
H> execve("/bin/mkdir", ["mkdir", "two"], [/* 18 vars */]) = 0
H> brk(0) = 0x9342000
H> access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file ...
H> mmap2(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS...
H> access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such file ...
H> open("/etc/ld.so.cache", O_RDONLY) = 3
H> fstat64(3, {st_mode=S_IFREG|0644, st_size=41270, ...}) = 0
H> mmap2(NULL, 41270, PROT_READ, MAP_PRIVATE, 3, 0) = 0xb78be000
H> close(3) = 0
H> access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file ...
H> open("/lib/libselinux.so.1", O_RDONLY) = 3
H> ...
I installed a more recent version of GNU coreutils from source, and
compared /bin/mkdir to /usr/local/bin/mkdir. I found /bin/mkdir gives
output like yours, but the /usr/local output is much shorter and doesn't
include any references to selinux:
me% strace -r /usr/local/bin/mkdir three
0.00 execve("/usr/local/bin/mkdir", ["mkdir", "three"], ...
0.000480 brk(0) = 0x196f2000
0.76 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|...
0.85 uname({sys="Linux", node="cmswbk002", ...}) = 0
0.000139 access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such ...
0.99 open("/etc/ld.so.cache", O_RDONLY) = 3
0.66 fstat(3, {st_mode=S_IFREG|0644, st_size=67370, ...}) = 0
0.000103 mmap(NULL, 67370, PROT_READ, MAP_PRIVATE, 3, 0) = ...
0.55 close(3) = 0
0.71 open("/lib64/libc.so.6", O_RDONLY) = 3
0.65 read(3, "\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0\...
0.73 fstat(3, {st_mode=S_IFREG|0755, st_size=1717800, ...}) =...
0.93 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_A...
0.71 mmap(0x32bce0, 3498328, PROT_READ|PROT_EXEC, MAP_PRI...
0.79 mprotect(0x32bcf4e000, 2093056, PROT_NONE) = 0
0.63 mmap(0x32bd14d000, 20480, PROT_READ|PROT_WRITE, MAP_PRIV...
0.78 mmap(0x32bd152000, 16728, PROT_READ|PROT_WRITE, MAP_PRIV...
0.62 close(3) = 0
0.71 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_A...
0.60 arch_prctl(ARCH_SET_FS, 0x2ae92dd5f230) = 0
0.000143 mprotect(0x32bd14d000, 16384, PROT_READ) = 0
0.69 mprotect(0x32bc01b000, 4096, PROT_READ) = 0
0.56 munmap(0x2ae92dd4d000, 67370) = 0
0.000141 mkdir("three", 0777) = 0
0.000148 close(1) = 0
0.60 close(2) = 0
0.62 exit_group(0) = ?
"-r" prints a relative timestamp upon entry to each system call.
H> This run was fast (less than 1 second). Unfortunately I do not know
H> when mkdir takes long and can't reproduce it today, so I can't tell
H> whether a strace of a slow mkdir is different.
Try this version of mkdir earlier in your PATH until you find out what
the delay is.
you% mkdir /tmp/mkdir
you% chmod 1777 /tmp/mkdir
you% cat /usr/local/bin/mkdir
#!/bin/sh
tfile=$(date '+%Y%m%d%H%M%S')
exec /usr/bin/strace -r -o /tmp/mkdir/$tfile /bin/mkdir ${1+"$@"}
exit 1
--
Karl Vogel I don't speak for the USAF or my company
NyQuil -The stuffy, sneezy, why-the-hell-is-the-room-spinning medicine.
--seen on T-shirt
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110522020507.468fab...@kev.msw.wpafb.af.mil
Re: first mkdir takes a long time (on ext3)
>> On 19/05/11 17:01, Hartmut Niemann wrote: H> It often takes very long time (20s) to mkdir on an ext3 drive What does "strace mkdir /some/directory" show? -- Karl Vogel I don't speak for the USAF or my company Q: What lies on the bottom of the ocean and twitches? A: A nervous wreck. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110520040714.1196ab...@kev.msw.wpafb.af.mil
Re: Format ext3 hard drives
>> On Tuesday 22 March 2011 02:42:36 pm Dan wrote: D> I am using the netinst to install Debian. I have one hard drive of 160GB D> and 2 hard drives of 2TB. Each hard drive has a ext3 partition for the D> whole drive. I used ext3 instead of ext4, because that is the default D> value in Squeeze. The netinst is creating the ext3 partitions but it is D> taking for ever. >> On Tue, 22 Mar 2011 15:19:12 -0800, Greg Madden said: G> Not sure about the initial format but all subsequent fsck take lots of G> time with ext3, esp 2 TB. If you're really set on ext3, specifying larger files and fewer inodes will definitely shorten the partition build and fsck times. I have some Seagate 1.5Tb drives, and I decided on two ~700Gb partitions instead of a single giant one. You might have to fool around with fdisk to get the sizes just the way you like. Details are below. -- Karl Vogel I don't speak for the USAF or my company Q: What's black and tan and looks good on a lawyer? A: A Rottweiler.--John J. Irvine --- Disk use: me% df Filesystem 1M-blocks Used AvailableUse% Mounted on /dev/sdb1 130 6 117 5% /boot /dev/sdb219501 445 18080 3% /root /dev/sdb5 7805 418 6997 6% /var /dev/sdb6 699594235065450509 35% /space1 /dev/sdb7 699601161094524487 24% /space2 me% df -i Filesystem Inodes IUsed IFree IUse% Mounted on /dev/sdb1 34136 11 34125 1% /boot /dev/sdb2 313344 11 31 1% /root /dev/sdb5 126976 11 126965 1% /var /dev/sdb611216896 169292 11047604 2% /space1 /dev/sdb711216896 64416 11152480 1% /space2 Partitions: # partition table of /dev/sdb unit: sectors /dev/sdb1 : start= 63, size= 273042, Id=83, bootable /dev/sdb2 : start= 273105, size= 40017915, Id=83 /dev/sdb3 : start= 40291020, size= 2008125, Id=82 /dev/sdb4 : start= 42299145, size=2887362450, Id= 5 /dev/sdb5 : start= 42299208, size= 16016742, Id=83 /dev/sdb6 : start= 58316013, size=1435664727, Id=83 /dev/sdb7 : start=1493980803, size=1435680792, Id=83 This script ran in about 25 minutes on a CentOS system: #!/bin/ksh -x # make filesystems with fewer inodes, larger files. export PATH=/sbin:/bin:/usr/bin date; mkfs.ext3 /dev/sdb1 date; mkfs.ext3 -J size=400 -i 65536 /dev/sdb2 date; mkswap -L SWAP-sdb3 /dev/sdb3 date; mkfs.ext3 -J size=400 -i 65536 /dev/sdb5 date; mkfs.ext3 -J size=400 -i 65536 -m 2 /dev/sdb6 date; mkfs.ext3 -J size=400 -i 65536 -m 2 /dev/sdb7 exit 0 Using fewer inodes added a total of nearly 80 Gb of available space on /space1 and /space2, compared to one huge partition with default inode setup. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110323012350.3a7f2b...@kev.msw.wpafb.af.mil
Re: putting "/tmp" to memory help
>> Dne, 23. 01. 2011 17:19:41 je Pascal Hambourg napisal(a): P> Tmpfs is not a RAM disk (RAM-based block device), it is a filesystem in P> virtual memory. >> On Sun, 23 Jan 2011 20:39:36 +0100, >> Klistvud said: K> Didn't know that. Damn clever. I stand corrected. http://landley.net/writing/rootfs-intro.html has a nice, short description of ramdisk vs. ramfs. -- Karl Vogel I don't speak for the USAF or my company The San Francisco Cable cars are the only mobile National Monuments. --odd but true -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110124004048.02cc7b...@kev.msw.wpafb.af.mil
Re: need help making shell script use two CPUs/cores
>> On Sun, 09 Jan 2011 10:05:43 -0600, >> Stan Hoeppner said: S> #! /bin/sh S> for k in $(ls *.JPG); do convert $k -resize 1024 $k; done Someone was ragging on you to let the shell do the file expansion. I like your way better because most scripting shells aren't smart enough to realize that when there aren't any .JPG files, I don't want the script to echo '*.JPG' as if that's actually useful. S> I use the above script to batch re-size digital camera photos after I S> dump them to my web server. It takes a very long time with lots of new S> photos as the server is fairly old, even though it is a 2-way SMP, S> because the script only runs one convert process at a time serially, S> only taking advantage of one CPU. First things first: are you absolutely certain that running two parallel jobs will exercise both CPUs? I've seen SMP systems that don't exactly live up to truth-in-advertising. If you stuff two "convert" jobs in the background and then run "top" (or the moral equivalent) do you SEE both CPUs being worked? Second: do you have "taskset" installed? If the work isn't being divided up the way you like, you can bind a process to a desired core: http://planet.admon.org/how-to-bind-a-certain-process-to-specified-core/ And last: if you're not using something like LVM, can you do anything to make sure you're not hitting the same disk? If all your new photos are on the same drive, any CPU savings you get from parallel processing will probably be erased by disk contention. Better yet, do you have enough memory to do the processing on a RAM-backed filesystem? -- Karl Vogel I don't speak for the USAF or my company If you're searching for the cause of a ghastly noise and find out that it's not the cat, leave the area immediately. --how to survive a horror movie -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/2011011500.46f09b...@kev.msw.wpafb.af.mil
Re: recursively find duplicate filenames
>> On Wed, 5 Jan 2011 03:15:01 -0800 (PST), >> S Mathias said: S> find duplicate filenames in a folder recursively? how? It's a (pretty ugly) one-liner if you have something to reverse strings in a file. Put this in (say) /usr/local/bin/rev: #!/usr/bin/perl -n chomp; print +(scalar reverse), "\n"; exit(0); And then: you% cd /some/place you% find . -print | rev | cut -f1 -d/ | rev | sort | uniq -c | \ expand | grep -v ' 1 ' Or you could replace the uniq-expand-grep with some awk weirdness if you're into that. -- Karl Vogel I don't speak for the USAF or my company If men ruled the world #18: It would be perfectly legal to steal a sports car, as long as you returned it the following day with a full tank of gas. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110106024045.89094b...@kev.msw.wpafb.af.mil
Re: recursively count the words occurrence in the text files
>> On Thu, 30 Dec 2010 10:34:59 -0800 (PST),
>> S Mathias said:
S> ...recursively count the words occurrence in the text files.
This assumes words consist of alphanumeric characters only. If that's
not the case, you'll need to change the "tr/" line in the script.
me% head */*
==> one/asf.txt <==
word1 word2 word3
==> one/asfcxv.txt <==
word2 word4, word5
==> one/dsgsdg.txt <==
word1. word2
==> three/qwerbdsg.txt <==
word7, word8 word9 word10
==> three/weberg.txt <==
word4 word3
==> three/werdf.txt <==
asdf, word2
==> two/ergd.txt <==
word6
==> two/sdgsddsf.txt <==
word6, word3
me% cat words
#!/usr/bin/perl -w
use strict;
my %words;
while (<>) {
tr/a-zA-Z0-9/ /cs;
foreach my $w (split) {
$words{$w}++;
}
}
foreach (sort keys %words) {
print "$_ $words{$_}\n";
}
exit(0);
me% ./words */*
asdf 1
word1 2
word10 1
word2 4
word3 3
word4 2
word5 1
word6 2
word7 1
word8 1
word9 1
--
Karl Vogel I don't speak for the USAF or my company
Never take a beer to a job interview. --Martha Stewart's tips for Rednecks
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20101230213326.cc245b...@kev.msw.wpafb.af.mil
Re: Extremely large level 1 backups with dump
>> In an earlier message, I said: K> This box has ~630,000 files using 640 Gbytes, but not many files change K> hourly. >> On Mon, 6 Dec 2010 21:33:01 -0700, Bob Proulx said: B> Note that you must have sufficient ram to hold the inodes in buffer cache. B> Otherwise I would guess that it would be hugely slower due to the need to B> read the disk while reading directories. But if there is sufficient ram B> for filesystem buffer cache then it will be operating at memory speeds. B> For anyone trying to recreate this goodness but wondering why they aren't B> seeing it then check that the buffer cache is sufficiently large. Here are some machine specifics for perspective. It's an IBM x3400, 2 Xeon 2GHz CPUs, 4Gb memory running RedHat. It has 8 WD4000KS 400-Gb drives, 16Mb buffersize, 300 MBps transfer rate, Serial ATA-300, 7200 rpm. me% free total used free sharedbuffers cached Mem: 19439481576708 367240 0 295304 947900 -/+ buffers/cache: 3335041610444 Swap: 20964723362096136 me% cat /proc/meminfo MemTotal: 1943948 kB MemFree:689696 kB Buffers:394412 kB Cached: 461864 kB SwapCached: 4 kB Active: 500328 kB Inactive: 419836 kB HighTotal: 1179008 kB HighFree: 681396 kB LowTotal: 764940 kB LowFree: 8300 kB SwapTotal: 2096472 kB SwapFree: 2096136 kB Dirty: 512 kB Writeback: 0 kB AnonPages: 63836 kB Mapped: 24212 kB Slab: 321088 kB PageTables: 4436 kB NFS_Unstable:0 kB Bounce: 0 kB CommitLimit: 3068444 kB Committed_AS: 196428 kB VmallocTotal: 114680 kB VmallocUsed: 3304 kB VmallocChunk: 111332 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 Hugepagesize: 4096 kB The default setup for read-ahead on the drives was 256 blocks. I did some testing and found my sweet spot was 16k: root# blockdev --getra /dev/sda 256 root# blockdev --setra 16384 /dev/sda The default device scheduler was Completely Fair Queue: root# cat /sys/block/sda/queue/scheduler noop anticipatory deadline [cfq] CFQ is 1% faster than elevator for a single user. I found a web link claiming that, in multi-user tests with 4 users, deadline had 20% better performance. At boot time: echo "deadline" > /sys/block/sda/queue/scheduler All filesystems are ext3, mounted like so: rw,nodev,noatime,nodiratime,data=journal They have the largest journal size possible, 400 Mb. In /etc/sysctl.conf: # Better write performance, avoids unnecessary paging. vm.swappiness = 10 Ran a time test concurrently with 49 active Samba sessions: date touch -d yesterday /tmp/TIME # "list" holds 8 340-Gb filesystems, from 42-67% full. find $list -newer /tmp/TIME -print | wc -l date Results: Tue Dec 7 13:34:05 EST 2010 541 Tue Dec 7 13:39:40 EST 2010 -- Karl Vogel I don't speak for the USAF or my company Adam was a Canadian. Nobody but a Canadian would stand beside a naked woman and worry about an apple. --Gord Favelle -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101207191114.c5577b...@kev.msw.wpafb.af.mil
Re: Extremely large level 1 backups with dump
>> On Sun, 5 Dec 2010 19:15:25 -0800, >> Peter Tenenbaum said: P> After having some difficulty getting rsync to do exactly what I want, P> I've become convinced to try rsnapshot. I'll let you know how it goes. I'm interested in seeing what kind of grief you're getting from rsync. I've had to argue with it in the past; feel free to reply privately if you'd rather. Don't rule out dumb and strong, it works great for me. On one of my servers, I do incremental backups hourly like so: cd /some/where find . -newer /some/timestamp -print > LIST ... [dork around with LIST] pax -rwdv -pe /else/where < LIST touch /some/timestamp This box has ~630,000 files using 640 Gbytes, but not many files change in an hour. The find command above rarely takes more than a few minutes. If you leave out the pax, you'll have a list of recently-changed files which you could feed to rsync via the --files-from=LIST argument. -- Karl Vogel I don't speak for the USAF or my company I would rather be exposed to the inconveniences attending too much liberty than to those attending too small a degree of it.--Thomas Jefferson -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101207033502.e12b7b...@kev.msw.wpafb.af.mil
Re: scripting question: to parse data with varname=value pattern the easiest way?
>> On Mon, 01 Nov 2010 15:49:01 +0800,
>> Zhang Weiwu said:
Z> A program output is like this:
Z> result_01: a="23" b="288" c="A_string" ac="34"
Z> result_02: a="23" b="28" c="A_string_too" dc="3"
Z>
Z> I am writing a script to output values of b if b is in the result set.
If your data is trustworthy and follows shell assignment rules like the
example above, you can abuse sh/ksh/bash:
#!/bin/bash
sample='result_01: a="23" b="288" c="A_string" ac="34"
result_02: a="23" c="A_string_too" dc="3"
result_03: a="23" b="28" c="A_string_too" dc="3"'
echo "$sample" | while read str
do
unset b # or whatever you're looking for
set $str # will fail horribly if $str is empty
result=$1
shift
eval "$@"
test "$b" && echo "$result $b"
done
exit 0
On the other hand, if someone sneaks something like
result_04: dc="3" rm /something/valuable
into your program output, you'll get a nasty surprise.
--
Karl Vogel I don't speak for the USAF or my company
Most users (myself included) spend most of their time in front of a computer
in a kind of fuzzy autopilot mode, and anything that creates ripples on that
placid lake of unawareness is going to be noticed as a disproportionately
significant problem. --David Harris, creator of Pegasus Mail
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20101101210436.50fccb...@kev.msw.wpafb.af.mil
Re: see what files are touched by a program
>> On Sun, 17 Oct 2010 21:57:36 +0200,
>> Steven said:
S> Can anyone point me to a program that will let me know what files a
S> certain application touches? More importantly, writes to. Something
S> like "filestouched vi filename" which would then report 'filename' as
S> being used.
"strace" will do what you want, but it's awkward to use if this app is
being called by something else. The easiest way might be to use a
small script as a placeholder for the application:
you% mv /path/to/app /path/to/app.bin
you% cat /path/to/app
#!/bin/sh
strace -e trace=open -o /tmp/x$$ /path/to/app.bin ${1+"$@"}
grep -v RDONLY /tmp/x$$ > /tmp/app$$
exec rm /tmp/x$$
exit 1
Running this using "vim" for the command and "stuff" for the filename
gave me this trace in /tmp/app21656:
open("/tmp/stuff.swp", O_RDWR|O_CREAT|O_EXCL|O_LARGEFILE, 0600) = 4
open("4913", O_WRONLY|O_CREAT|O_EXCL|O_LARGEFILE, 0100644) = 3
open("stuff", O_WRONLY|O_CREAT|O_TRUNC|O_LARGEFILE, 0644) = 3
open("/me/.viminfo.tmp", O_WRONLY|O_CREAT|O_EXCL|O_LARGEFILE, 0600) = 5
--
Karl Vogel I don't speak for the USAF or my company
He'd make a lovely corpse. --Charles Dickens
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20101020014828.c184bb...@kev.msw.wpafb.af.mil
Re: how to configure gcc
>> On Thu, 30 Sep 2010 09:20:36 +0530,
>> Anand Sivaram said:
AS> On Thu, Sep 30, 2010 at 09:12, Boyd Stephen Smith Jr. wrote:
B> On Wednesday 29 September 2010 15:18:46 abdelkader belahcene wrote:
A> yes but normally gcc should be configured, where is the config file,
B> It's in the same place as the config file for ls, cp, and rm. (There
B> isn't one.)
This is where the BSD version of make has an advantage over GNU make.
It checks for system-wide defaults in the file /etc/make.conf:
me% uname -sr
FreeBSD 6.2-RELEASE
me% cat /etc/make.conf
CPUTYPE?=pentium4
CFLAGS= -O2 -pipe -funroll-loops
COPTFLAGS= -O -pipe
AS> I think using an alias 'gcc -lm' is not a good idea. It may be looking
AS> easier for this particular C file, but '-lm' would be linked always for
AS> even in the case of hello_world.c which is really unnecessary.
The additional linking won't do much besides take an extra millisecond
of time and add around 25 bytes to your binary; that's what I saw when
comparing on a Deadrat Enterprise system. Build the executable both
ways, run "nm" and chop the first 10 characters to see for yourself.
The results should be identical.
You're right about the alias -- it works fine right up to the part where
someone changes their login shell or alias setup. It's safer to use a
script as a front-end, so you can check for environment variables in a
global or user-specific config file:
#!/bin/sh
# front-end for make with preset preferences.
export PATH=/usr/local/bin:/bin:/sbin:/usr/sbin:/usr/bin
cbase='make.conf'
for cfg in /etc/$cbase $HOME/etc/$cbase; do
test -f "$cfg" && . $cfg
done
exec /usr/bin/make ${1+"$@"} # Use full path just to be safe...
exit 1
Under Linux, the config file looks like this:
me% cat $HOME/etc/make.conf
# Default values for make.
export CC="gcc"
export CFLAGS="-O2 -pipe -funroll-loops"
export COPTFLAGS="-O -pipe"
Comparison:
me% cat Makefile
hello: hello.o
hello.o: hello.c
me% cat hello.c
#include
#include
main()
{
printf("hello, world\n");
#ifdef unix
printf("unix defined\n");
#endif
exit(0);
}
me% make
cc -c -o hello.o hello.c
cc hello.o -o hello
me% ./make # using the frontend script
gcc -O2 -pipe -funroll-loops -c -o hello.o hello.c
gcc hello.o -o hello
--
Karl Vogel I don't speak for the USAF or my company
Beauty is only a light switch away.
--Perkins Library, Duke University, Durham, North Carolina
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100930192027.79c5ab...@kev.msw.wpafb.af.mil
Re: root can't sudo
>> On Tue, 28 Sep 2010 09:42:40 -0400 (EDT), >> Stephen Powell said: S> I don't speak for the OP, but my guess is that the OP has a script that S> he wants to be able to run either as his non-superuser self or as root. Easy. This preserves arguments including spaces: #!/bin/sh PATH=/usr/local/bin:/bin:/usr/bin; export PATH test "`id -u`" -gt 0 && exec sudo $0 "$@" whoami for arg in "$@"; do echo "[$arg]" done exit 0 -- Karl Vogel I don't speak for the USAF or my company EXCUSE FOR GETTING TO WORK LATE #7: The dog ate my car keys. We're going to hitchhike to the vet. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100928185441.e2d39b...@kev.msw.wpafb.af.mil
Re: rsync issue
>> On Sun, 12 Sep 2010 18:30:59 +0200, >> Rodolfo Medina said: R> I wish to copy all my home directory into another machine with the R> --delete option, but: R> 1) I don't want hidden files, i.e.: `.*' to be copied; R> 2) there are some symlinks, beginning with `.', that I want to be copied. Instead of copying everything and deleting what you don't want, why not make a list of the files you want copied and have rsync use that instead? Something like this: list=/tmp/list$$ # or use mktemp cd $HOME # Ignore hidden files and "." directory... find . -print | egrep -v '^(./\.|\.$)' > $list # ...but keep symlinks starting with a dot. find . -type l -name '.*' -print >> $list rsync --files-from=$list [other rsync options, destination, etc] rm $list You can add any other filtering you might want after the last "find", you *know* exactly what files are being synced, and you can store that list someplace in case there's ever a question of why something did (or didn't) get copied. -- Karl Vogel I don't speak for the USAF or my company I have never hated a man enough to give his diamonds back. --Zsa Zsa Gabor -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100913011305.475abb...@kev.msw.wpafb.af.mil
Re: Dumping terminal contents to a file
>> On Wed, 08 Sep 2010 12:52:05 -0400 (EDT), Alois Mahdal wrote:
A> I wonder if there's a simple and universal way to dump output of
A> terminal to a file, particularly for demonstration purposes.
>> On Wed, 8 Sep 2010 13:35:30 -0400 (EDT), Stephen Powell
>> said:
S> Well, for recording a shell session, there is "script". And to play it
S> back later, there is "scriptreplay". [...] I don't know if that's what
S> you want, though. script puts *everything* in the capture file,
S> including linefeeds, backspaces, ANSI escape sequences, etc. For
S> example, colorized output from "ls" results in ANSI escape sequences.
I use a wrapper called "saveon" which tries to correct the worst results
from "script" putting everything in the capture file:
#!/bin/sh
PATH=/bin:/usr/bin:/usr/local/bin
SAVEON=1
export PATH SAVEON
case "`whoami`" in
root) prompt='root#' ;;
*)prompt='me%' ;;
esac
export PS1="
$prompt "
tmp=`mktemp /tmp/$tag.XX` || {
echo "$tag: mktemp failed" >&2; exit 1
}
trap "rm -f $tmp; exit 0" 0 1 2 3 15
sedscr='
s/^.*script-is-done://
s/^.*script-started-on://
/Script started on /d
/^me% *$/d
/^me% exit$/d
/^root# *$/d
'
echo "script-started-on:`date '+%a %b %d %T %Y'`" > $tmp
script -c /bin/sh -a -q $tmp
echo "script-is-done:`date '+%a %b %d %T %Y'`" >> $tmp
col -b < $tmp | sed -e "$sedscr" > typescript
exit 0
* The date commands give me a consistently-formatted start and finish
time at the top and bottom of "typescript".
* The newline in the prompt cleanly separates commands from output.
* "col -b" gets rid of at least some special characters.
* The SAVEON environment variable lets me set programs to show colors or
mess with terminal settings only when it makes sense to do so. For
example, my "dir" script looks like this:
#!/bin/sh
case "$SAVEON" in
"") opt='--color=auto' ;;
*) opt='' ;;
esac
unset BLOCK_SIZE # throws off the results.
exec ls -lF $opt ${1+"$@"}
--
Karl Vogel I don't speak for the USAF or my company
You do not need a parachute to skydive. You only need a parachute to
skydive twice.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100909020843.43f0ab...@kev.msw.wpafb.af.mil
Re: match across line using grep
>> On Mon, 02 Aug 2010 14:56:45 +0800,
>> Zhang Weiwu said:
Z> I'm grepping a bunch of files each have a segment code that executes a
Z> SQL. My problem is that the query spans across several lines and I
Z> can't seem to make grep honor (?s) for that.
Perl Is Our Friend. Here's some text to search:
me% cat -n sample
1 Message-ID: <4c566c2d.7000...@realss.com>
2 Date: Mon, 02 Aug 2010 14:56:45 +0800
3 From: Zhang Weiwu
4 Organization: Real Softservice
5 Status: RO
6 Content-Length: 2410
7 Lines: 80
8
9 I'm grepping a bunch of files each have a segment code
10 that executes a SQL. Both selects should match:
11
12 select * from table1 where id=1;
13
14 select * from table2 where id=1
15 and name='foo'";
16
17 I tried to use -z parameter for grep, which the manual says
18 would make grep not treating \n as line terminator. But
19 it doesn't work neither. A simple test shows I might have
20 misunderstood the use of -z:
The script below my signature gives these results:
me% ./pgrep 'select.*from.*;' sample
[sample:12] select * from table1 where id=1;
matched >>select * from table1 where id=1;<<
[sample:15] select * from table2 where id=1
and name='foo'";
matched >>select * from table2 where id=1
and name='foo'";<<
"[sample:15]" means the match happened at or before line 15 in file
"sample". You can pass multiple files on the command line. The
"matched" stuff is there in case there's some distracting text on the
line besides the select statement. If you want matching to be
case-sensitive, change "si:" to "s:" on the line where $pattern is set.
--
Karl Vogel I don't speak for the USAF or my company
---
#!/usr/bin/perl -w
# Taken from perl-grep3.pl in "Mastering Perl"
use strict;
# Get the desired pattern, make newlines match '.' and ignore case.
my $pattern = shift @ARGV || die "I need a pattern\n";
$pattern = '(?si:' . $pattern . ')';
# Make sure pattern works.
my $regex = eval { qr/$pattern/ };
die "Check your pattern! $@" if $@;
# Use paragraph mode to handle newlines.
$/ = "";
my $line = 0;
while (<>) {
$line += tr/\n/\n/;
chomp;
print "[$ARGV:", $line-1, "] $_\n\t\tmatched >>$&<<\n\n" if m/$regex/;
$line = 0 if eof(ARGV);# reset counter for a new file.
}
exit(0);
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100802214030.a21b6b...@kev.msw.wpafb.af.mil
Re: Shell Expansion in Bourne Shell Script Question
>> On Thu, 29 Jul 2010 01:04:27 -, >> Cameron Hutchison said: C>find $MAGDIR -iname '*.zip' -print0 | xargs -0 some-command C> -iname matches names case insensitively. Since you then dont need grep, C> you also dont need tr0. I need to think before posting. I didn't mention that I have FreeBSD, Linux, and Solaris boxes, and unfortunately I can't guarantee the same access to GNU find. I can install xargs if the system version doesn't recognize the "-0" option, so I usually end up scripting for the lowest common denominator. -- Karl Vogel I don't speak for the USAF or my company Cute Celebrity Couple Names #6: Lance Armstrong + Ivanka Trump = Armstump -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100729023745.d8a58b...@kev.msw.wpafb.af.mil
Re: Shell Expansion in Bourne Shell Script Question
>> On Wed, 28 Jul 2010 23:58:11 +0200,
>> Mart Frauenlob said:
M> One might be better of with some like this:
M> find /DIR -regextype posix-egrep -regex '.*\.(zip|ZIP)' -exec \
M> some_command {} +
If the filelist is potentially too big for the max argument list on the
system, I would do something like this:
find $MAGDIR -print | grep -i '\.zip$' | tr0 | xargs -0 some command
We have a ton of filenames with spaces and other crap, so "tr0" just
replaces newlines with nulls:
#!/bin/sh
PATH=/bin:/usr/bin
export PATH
exec tr '\012' '\000'
exit 1
--
Karl Vogel I don't speak for the USAF or my company
Q. Why are New Yorkers always depressed?
A. The light at the end of the tunnel is New Jersey. --Fredrick Homan
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100729003203.9667cb...@kev.msw.wpafb.af.mil
Re: Shell Expansion in Bourne Shell Script Question
>> On 28.07.2010 14:42, Jochen Schulz wrote: J> I think you meant to write J> for MAGFILE in `ls $MAGDIR/*.[Zz][Ii][Pp]` J> Another hint: you don't need 'ls' for your case at all. I'd recommend keeping the "ls". Try your script when MAGDIR doesn't have any zipfiles, and MAGFILE will hold the expanded value of $MAGDIR with the string "*.[Zz][Ii][Pp]" appended. Bash, sh, and ksh all do that for some dopey reason, which is why I would do for file in $(ls $MAGDIR/*.[Zz][Ii][Pp] 2> /dev/null); do ... if there's any chance that you won't find any files. -- Karl Vogel I don't speak for the USAF or my company Love and respect for the great American experiment in free government does not appear out of thin air. --Larry P. Arnn -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100728180522.353a1b...@kev.msw.wpafb.af.mil
Re: Monitoring tools to use on an account
>> On Tue, 27 Jul 2010 17:53:40 +0100, AG wrote: A> I'm facing a bit of a delicate issue: I have created an account on my A> machine for someone staying with us, and I have strong suspicions that A> he is engaging in on-line behaviour that he is not supposed to be doing. A> Can anyone recommend a tool thatb I can install, that can monitor his A> on-line activity - specifically sites he visits and how much time he A> spends on them? >> On Wed, 28 Jul 2010 11:05:30 + (UTC), said: C> If he has nothing to hide, all the steps will be tracked by the browser C> history and cache files. Also, "/tmp" is a good bucket for holding C> "shared secrets" (recent files, etc...). The problem is if he does have something to hide that the OP might be held liable for. AG, if you're worried about browser activity, can you install squid on your system and change his proxy setting accordingly? This way he leaves a trace even if he sanitizes his browser cache, assuming he doesn't have root privileges. Another possibility - running tcpdump or the moral equivalent and checking the packet dumps periodically for anything hinky. This way you catch any bad network activity, not just the browser. Something like this at boot to avoid filling your entire drive: k=1 while true; do out=/some/dir/dump.$k # /some/dir owned by you, mode 700 tcpdump -c 50 -w $out # season to taste # check the dump for anything suspicious, remove it if clean tcpdump -r $out ... some filter here ... || rm $out k=$((k+1)) done You might also change the permissions on "ps" so he can't see "tcpdump" or any other steps you might take. -- Karl Vogel I don't speak for the USAF or my company If you can't be kind, at least have the decency to be vague. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100728175429.7333ab...@kev.msw.wpafb.af.mil
Re: Sed or awk: remove a line from a file
Here are some good pages on sed and awk one-liners: http://www.catonmat.net/blog/sed-one-liners-explained-part-one/ Famous Sed One-Liners Explained, Part I: File Spacing, Numbering and Text Conversion and Substitution Wed, 8 Oct 2008 03:00:00 -0400 http://www.catonmat.net/blog/sed-one-liners-explained-part-two/ Famous Sed One-Liners Explained, Part II: Selective Printing of Certain Lines Sun, 21 Dec 2008 14:45:00 -0500 http://www.catonmat.net/blog/sed-one-liners-explained-part-three/ Famous Sed One-Liners Explained, Part III: Selective Deletion of Certain Lines and Special Applications Wed, 14 Jan 2009 17:10:00 -0500 http://www.catonmat.net/blog/awk-one-liners-explained-part-one/ Famous Awk One-Liners Explained, Part I: File Spacing, Numbering and Calculations Sun, 28 Sep 2008 01:00:00 -0400 http://www.catonmat.net/blog/awk-one-liners-explained-part-two/ Famous Awk One-Liners Explained, Part II: Text Conversion and Substitution Sat, 13 Dec 2008 04:25:00 -0500 http://www.catonmat.net/blog/awk-one-liners-explained-part-three/ Famous Awk One-Liners Explained, Part III: Selective Printing and Deleting of Certain Lines Mon, 5 Jan 2009 10:00:00 -0500 http://www.catonmat.net/blog/update-on-famous-awk-one-liners-explained/ Update on Famous Awk One-Liners Explained: String and Array Creation Mon, 9 Feb 2009 10:25:00 -0500 -- Karl Vogel I don't speak for the USAF or my company If walking is so good for you, then why does my mailman look like Jabba the Hut? --bumper sticker -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100603003304.bb460b...@bsd118.wpafb.af.mil
Re: monitoring internet availability and sending sms alert?
http://www.hcst.net/~vogelke/src/ishostup/ These scripts are good for exactly one thing: sending me (or someone) a message if a given piece of hardware at a given IP address stops responding for more than a minute or so. They rely on fping and regular email. -- Karl Vogel I don't speak for the USAF or my company He kept saying I didn't listen to him... or something. --Tritchen Smith -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100517014401.a0ca8b...@bsd118.wpafb.af.mil
Re: bash scripting question
Here's something I modified as part of a benchmark script called "fdtree".
--
Karl Vogel I don't speak for the USAF or my company
Dijkstra probably hates me. --Linus Torvalds, in kernel/sched.c
#!/bin/bash
# How to use xdate/xtime/persec:
#
# START=$(date "+%s")
# count=10
# xdate
# tin=$(xtime)
#
# # do something time-consuming $count times...
# tout=$(xtime)
# set $(persec $tin $tout $count); ttot=$1; results=$2
# xdate
# echo "TIME IN, OUT, TOTAL = "$tin, $tout, $ttot
# echo -e "\tWork per second = " $results
PATH=/usr/local/bin:/bin:/usr/bin
export PATH
function xdate
# Display the date in this form: Thu, 10 Sep 2009 21:22:06.494
{
set $(date "+%a, %d %b %Y %T %N")
ms=$(echo $6 | cut -c1-3)
echo "$1 $2 $3 $4 $5.$ms"
}
function xtime
# Display elapsed runtime to the millisecond.
{
set $(date "+%s %N")
sec=$(($1 - $START))
ms=$(echo $2 | cut -c1-3)
echo "$sec.$ms"
}
function persec
# args: start-second, finish-second, count-things
# returns elapsed time and things that happened per second
# to the millisecond.
{
start=$1
finish=$2
count=$3
if test "$finish" = "$start"; then
echo "0 0"
else
echo $(echo "scale=3; $finish-$start;
$count/($finish-$start)" | bc)
fi
}
function dbg
# debugging prints
{
test $DEBUG -gt 0 && echo -e "$@"
}
tmp=/tmp/t1$$
tmp2=/tmp/t2$$
xdate
START=$(date "+%s")
tin=$(xtime)
echo
for k in 1 2 3 4 5 6 7 8 9 10
do
# Read a short amount of random data.
dd if=/dev/random of=$tmp2 bs=1k count=1 2> /dev/null
# Duplicate it a bunch of times.
cat $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 > $tmp
cat $tmp $tmp $tmp $tmp $tmp $tmp $tmp $tmp > $tmp2
cat $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 > $tmp
cat $tmp $tmp $tmp $tmp $tmp $tmp $tmp $tmp > $tmp2
cat $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 $tmp2 > $tmp
rm $tmp2
# Generate a hash.
md5sum $tmp
ls -l $tmp
rm $tmp
done
echo
count=10
tout=$(xtime)
set $(persec $tin $tout $count); ttot=$1; results=$2
xdate
echo "TIME IN, OUT, TOTAL = "$tin, $tout, $ttot
echo -e "\tWork per second = " $results
exit 0
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100330012125.34bebb...@bsd118.wpafb.af.mil
