Re: NIS user audit?
On Mon, 2003-12-01 at 15:16, Gavin Hamill wrote: > I'm running NIS on woody with no problems at all in the auth / shadow > file, etc. but I'd like to be able to track when each user logs in and > out, (basically a computer-based timeclock). > > I don't see any trace of the NIS server logging such data, and running > ruserd on each diskless workstation and polling the 'logged in users' > lost seems a little inelegant. > > At this stage, I'm considering calling 'wget > http://10.0.0.254/cgi-bin/login.cgi?username=' as part of their X > startup conf.. (and similar for logout) Diskless workstations. so you probably are already running a centralized syslog? -- Tot ziens, Bart-Jan
Re: NIS user audit?
On Mon, 2003-12-01 at 15:16, Gavin Hamill wrote: > I'm running NIS on woody with no problems at all in the auth / shadow > file, etc. but I'd like to be able to track when each user logs in and > out, (basically a computer-based timeclock). > > I don't see any trace of the NIS server logging such data, and running > ruserd on each diskless workstation and polling the 'logged in users' > lost seems a little inelegant. > > At this stage, I'm considering calling 'wget > http://10.0.0.254/cgi-bin/login.cgi?username=' as part of their X > startup conf.. (and similar for logout) Diskless workstations. so you probably are already running a centralized syslog? -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How were the apache debs compiled ?
On Wed, 2003-04-16 at 17:36, Maarten van der Hoef wrote: > I wonder how the apache of apache..deb & apache-common..deb (woody) was > compiled, > the ./configure line ... etc. > > I want to create my own debian apache package, but I want it to > be compatible with the other packages in the woody-tree that > depends on apache, or where apache depends on other packages. apt-get source apache cd apache-x.x.xx-x $PAGER debian/rules This requires a deb-src line in /etc/apt/sources.list similar to the deb line you already have there. Make sure to run apt-get update after updating the sources.list file. -- Tot ziens, Bart-Jan
Re: UPS sharing
On Tue, 2003-03-11 at 08:47, Craig wrote: > I was wondering if its possible to have 2 Debian boxes sharing > a single UPS ? Its a 650Pro - made by APC. Besides connecting both their power supplies to that UPS ?? :) I'm using the package nut with my APC SmartUPS 1000 and it's running fine. There even is a winnut port of the client code. -- Tot ziens, Bart-Jan
Re: UPS sharing
On Tue, 2003-03-11 at 08:47, Craig wrote: > I was wondering if its possible to have 2 Debian boxes sharing > a single UPS ? Its a 650Pro - made by APC. Besides connecting both their power supplies to that UPS ?? :) I'm using the package nut with my APC SmartUPS 1000 and it's running fine. There even is a winnut port of the client code. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Apache / PHP4 config problem (newbee on PHP)
On wo, 2003-01-15 at 17:32, Gregoire Hostettler wrote: > I moved the script to its original location (/var/www) but I get a strange > behavior: > > - Using IE6, the browser asks me if I want to download the test.php script > (???) bartjan@trillian:~$ grep php4 /etc/apache/mime.types application/x-httpd-php4php4 I think that you don't have the correct mime type set for php files in the /etc/apache/mime.types or Apache is not using this file. > - If I rename the .php script to .html, it displays just the header, nothing > else, which is correct imho Yes. What you could do now is to add .html to the line in srm.conf (newer Apache versions have all options in httpd.conf) that already has: AddType application/x-httpd-php .php So that you get: AddType application/x-httpd-php .php .html This adds php-support for .html files in adition to .php files. > - If I try to access test.php via lynx, it just displays the html text. > There is a tmp html file created, althought > > It seems to me that the php4 engine never starts to generate pages... > > Any other idea ? Yes. Please do not top-post and learn to strip what you are quoting. > > On wo, 2003-01-15 at 13:50, Gregoire Hostettler wrote: > > > > > Got a (supposed) config problem with PHP4/Apache. > > [Wed Jan 15 14:29:55 2003] [notice] Apache/1.3.9 (Unix) Debian/GNU > PHP/4.0.3pl1 mod_perl/1.21_03-dev configured -- resuming normal operations > > [Wed Jan 15 14:29:55 2003] [notice] suEXEC mechanism enabled (wrapper: > /usr/lib/apache/suexec) > > [Wed Jan 15 14:31:49 2003] [error] (8)Exec format error: exec of > /usr/lib/cgi-bin/test.php failed > > [Wed Jan 15 14:31:49 2003] [error] [client 192.168.124.1] Premature end of > script headers: /usr/lib/cgi-bin/test.php > > > > You are placing html files in cgi-bin directories. Don't do that, > > because cgi-bin directories are reserved only for cgi scripts, not html > > or php files. Just place the .php file in /var/www and I think it should > > work. -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Apache / PHP4 config problem (newbee on PHP)
On wo, 2003-01-15 at 13:50, Gregoire Hostettler wrote: > Got a (supposed) config problem with PHP4/Apache. > > OS: Debian potato 2.2r7, kernel 2.2.19 > Apache 1.3.9 (Debian implementation) > PHP4 4.0.3pl1-0p > > Attached are both the httpd.conf and srm.conf files, along with the > access.log, error.log, php script involved and the output of apache -V > (apachev file). Only the error.log was enough: [Wed Jan 15 14:29:55 2003] [notice] Apache/1.3.9 (Unix) Debian/GNU PHP/4.0.3pl1 mod_perl/1.21_03-dev configured -- resuming normal operations [Wed Jan 15 14:29:55 2003] [notice] suEXEC mechanism enabled (wrapper: /usr/lib/apache/suexec) [Wed Jan 15 14:31:49 2003] [error] (8)Exec format error: exec of /usr/lib/cgi-bin/test.php failed [Wed Jan 15 14:31:49 2003] [error] [client 192.168.124.1] Premature end of script headers: /usr/lib/cgi-bin/test.php You are placing html files in cgi-bin directories. Don't do that, because cgi-bin directories are reserved only for cgi scripts, not html or php files. Just place the .php file in /var/www and I think it should work. -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How to avoid the screen timeout when no activity
On za, 2003-01-11 at 14:05, Michelle Konzack wrote: > Am 18:02 2003-01-07 -0800 hat Jeremy C. Reed geschrieben: > >On Wed, 8 Jan 2003, Samantha Scafe wrote: > > > >> How can avoid the timeout of the monitor. > > > >Look at setterm(1). > > > >Try: > > setterm -blank 0 > > This works only, if you have minimal one login after reboot... > Is there a possibility to get it without login ??? No need to login. Just add the command to a file in /etc/init.d/ and make sure that file is executable and is linked to the various /etc/rc?.d directories. See the update-rc.d manpage. -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: postfix oddities.... 220 *******
On vr, 2003-01-10 at 14:45, Roger Abrahamsson wrote: > Currently trying to set up a test postfix server with smtp auth, with > sasl and tls. Got everything working eventually, at least from > localhost. I can login, authenticate with plaintext etc. > However, when trying to access that server from a host not in > "mynetworks" it echos a funny 220 header like this: > > > Escape character is '^]'. > 220 PIX firewall ?? > if I then try a ehlo like: > > ehlo localhost > > I get: > 502 Error: command not implemented PIX firewalls don't know how to properly speak (e)smtp, and this causes errors. Best way to fix this is on the PIX itself (make sure you have the latest patches etc. and consider disabling the smtp fixup mode), but as a workaround you could disable esmtp on the postfix box, by having this in main.cf: smtp_always_send_ehlo = no smtp_never_send_ehlo = yes -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: backup
On Mon, 2002-12-16 at 19:50, Russell Coker wrote: > On Mon, 16 Dec 2002 19:36, C. R. Oldham wrote: > > We used to use Amanda before we got a tape library and began backing up > > all our servers into the library. We have a heterogenous network > > (Linux, Win2000 with Exchange, Oracle). My experience with Amanda is > > that it works pretty well if you > > are doing Unix clients only, but doing a heterogenous network requires > > lots of scripting. Plus Amanda doesn't drive tape changers/libraries > > out of the box. It needs some scripting to do that as well. > > I'm only interested in backing up Linux and Solaris servers (and Solaris is > optional) so lack of cross platform support is not a problem. This is no problem with Amanda. As was said, Amanda does not support Windows directly and uses smbclient via a gateway to do Windows backups and that's why our only Windows server has its own tapedrive. The only scripting I did was a wrapper that dumps the mysql databases and creates the LVM snapshots that will be backed up and ejects the tape when the backup was succesful (this way the human tape robot knows the backup was ok and next day's tape should be inserted). > How difficult is the scripting to change tapes? What exactly needs to be > done? Although I've never used a tapechanger myself, I think the only difficulty is that there are many types of tapechanges that are supported by Amanda and you'll need to find out which script you need for your changer. I haven't been on the [EMAIL PROTECTED] for a few years, but it is an excelent mailinglist. Amanda is great, but there are a few things it can't do: - append to tapes (it is a good thing that Amanda can't do this...) - spread a partition over multiple tapes. Another 'strange' behaviour of Amanda is that it does not do 'daily' and 'weekly' backups, it just spreads out all backups over its backup cycle and optimizes the tape usage. This gives better results than doing daily incrementals etc. but might be a hard thing to explain to your boss when he wants to do a full backup each friday. > Does anyone have some sample scripts? /usr/share/doc/amanda-common/docs/TAPE.CHANGERS.gz If you don't have amanda-common installed, or don't have the amanda source local, then you could use google to find a version, like on http://www.clug.in-chemnitz.de/vortraege/amanda/docs/TAPE.CHANGERS or somewhere else. -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Controlling IP traffic (e.g. per month)
On Mon, 2002-10-28 at 10:32, Dieter Heiliger wrote: > I am doing the administration for a server located at an ISP. We have a > certain amount of traffic per month included in the monthly fee. > Additional traffic will be paid extra. > > Now I want to get more control over the traffic that was already used > since the beginning of each month. My ideal configuration would be to > get an email when e.g. 80% of the traffic included in the monthly fee > got used already. When a new month begins the counter has to be set to > zero again. > > Is there an easy way to implement this? ipac (if you use ipchains) or ipac-ng (if using iptables). Then you can use the ipacsum command to generate a report. If you are only interested in the total traffic you could use something like: [EMAIL PROTECTED]:~#ipacsum -t "this month" --filter "all" --fixed-quantity '' IP accounting summary Host: spiderwebs / Time created: 2002/10/28 11:10:42 CE Data from 2002/10/01 01:00:00 CE to 2002/10/28 11:10:42 CE total in all: 4317454231 total out all : 4328291196 With a bit of perl or awk it's not difficult to calculate whether this is below or above your 80% threshold and decide if an email should be sent. -- Tot ziens, Bart-Jan Vrielink
Re: Controlling IP traffic (e.g. per month)
On Mon, 2002-10-28 at 10:32, Dieter Heiliger wrote: > I am doing the administration for a server located at an ISP. We have a > certain amount of traffic per month included in the monthly fee. > Additional traffic will be paid extra. > > Now I want to get more control over the traffic that was already used > since the beginning of each month. My ideal configuration would be to > get an email when e.g. 80% of the traffic included in the monthly fee > got used already. When a new month begins the counter has to be set to > zero again. > > Is there an easy way to implement this? ipac (if you use ipchains) or ipac-ng (if using iptables). Then you can use the ipacsum command to generate a report. If you are only interested in the total traffic you could use something like: root@spiderwebs:~#ipacsum -t "this month" --filter "all" --fixed-quantity '' IP accounting summary Host: spiderwebs / Time created: 2002/10/28 11:10:42 CE Data from 2002/10/01 01:00:00 CE to 2002/10/28 11:10:42 CE total in all: 4317454231 total out all : 4328291196 With a bit of perl or awk it's not difficult to calculate whether this is below or above your 80% threshold and decide if an email should be sent. -- Tot ziens, Bart-Jan Vrielink -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [postfix] virtual domains
On Wed, 2002-09-18 at 18:57, Davi Leal wrote: > How to reverse-map local users for out-bound mail, as in sendmail with > FEATURE(`genericstable'...) ?? > > Reference: http://www.sendmail.org/virtual-hosting.html See http://www.postfix.org/rewrite.html for a lot of ways to manipulate addresses with Postfix. I think a canonical map is what you'll want. See the canonical(5) manpage. -- Tot ziens, Bart-Jan
Re: [postfix] virtual domains
On Wed, 2002-09-18 at 18:57, Davi Leal wrote: > How to reverse-map local users for out-bound mail, as in sendmail with > FEATURE(`genericstable'...) ?? > > Reference: http://www.sendmail.org/virtual-hosting.html See http://www.postfix.org/rewrite.html for a lot of ways to manipulate addresses with Postfix. I think a canonical map is what you'll want. See the canonical(5) manpage. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Discrepencies between du and df.
On Fri, 2002-05-03 at 11:04, Fred Clausen wrote: > I am having a strange problem where the output of du and df do not > correlate. Here are the symptoms : > > $df /var > Filesystem 1k-blocks Used Available Use% Mounted on > /dev/hda10 606405596139 0 100% /var > > $du -s /var > 54157 /var > > Here we can see that du thinks there is only about 54M used but df reckons > there is over 596M used. Both are correct. The difference is that whenever an application has an open file, but the file is already deleted, then it is counted in the df output (because the space is certainly not free) but not in du (because it is not being used by a file). Just do a 'lsof | grep deleted' and I'm almost certain you'll see some large file in /var/log being held open by some daemon that wasn't restarted whenever it's logfile was rotated. Just restart that daemon (and don't forget to fix the logrotate scripts for that daemon and/or file a bugreport about it) -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: how to isolate iptables log ?
On Tue, 2002-04-30 at 14:21, Patrick Hsieh wrote: > Is it possible to seperate iptables log from /var/log/syslog? > I use some -j LOG stuff, but all log were put in syslog. How can I > collect them into different logfile, say /var/log/iptables.log ? man syslog.conf(5) Just make sure that the kern facility is not logged to /var/log/syslog. There are other syslog variants which are better at this, but I have no experience with them. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: avoid user direct accec *.html
On Mon, 2002-04-29 at 20:12, Patrick Hsieh wrote: > If I want to avoid user to directly access my .html files, say type the > complete url in the browser, is it possible? > > In PHP, I can check the HTTP_REFERER to make sure connections originates > from the same website. If the HTTP_REFERER is empty or not belongs to > the same website, I can redirect the client to another webpage. However, > when it comes to static .html or even .jpg files, is it possible to > configure apache to avoid that situation? We use mod_rewrite to block external websites from directly linking to images on one of our websites. Based on the situation, we either return a 403 Forbidden code or provide a watermarked version of the requested image. Please keep in mind the advice given by others on this list that you can't thrust user-supplied data. You can't use HTTP_REFERER to protect your data, but it's a great tool to stop external websites from stealing your bandwidth. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: File and directory permissions
On Mon, 2002-04-22 at 11:10, Martin Alfke wrote:
> What Martin sent in will work only if there are no white spaces
> in file- and directory-names.
>
> I would prefer this one:
>
> find [path] -type d -printf '"%p"\n' | xargs chmod 755
>
> (Please try first if %p is the correct term. Maybe
> %n is the correct substitution.)
Or use find [path] -type d -exec chmod 2775 {} \;
which is a little bit slower, but does not give you problems with spaces
in filenames.
Or use find [path] -type d -print0 | xargs -0 chmod 2775
because both find and xargs understand using null characters as
separators.
--
Tot ziens,
Bart-Jan
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Postfix Regular Expressions
On Sun, 2002-04-14 at 03:18, Gene Grimm wrote:
> Craig Sanders wrote:
> > \s and \t (and \d, \w, \b, and many others) are pcre (perl-compatible
> > regular expressions) which are an extension of standard regexps.
> >
> > > If the map is a regexp style map, then you just use a tab instead of
> > > \t.
> >
> > yep, and use "[[:space:]]" instead of "\s" - that matches spaces &
> > tabs, just like \s does.
>
> For pcre style expressions how would you represent the perl pattern
> /\s{15,}[a-zA-Z0-9]{2,8}$/? Is it exactly the same in pcre as it is in
> perl? How would you represent this in regexp style (for immediate spam
> rejection patterns)?
Well, as pcre stands for 'perl compatible re', I would guess it should
be exactly the same.
But the best answer you can get is testing itself by using postmap -q
with that regex pattern and a few test-lines.
I believe a regexp style pattern is almost the same, with only \s
replaced by [[:space:]]
See the pcre(7) and regex(7) manpages for the details.
--
Tot ziens,
Bart-Jan
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Postfix Regular Expressions
On Fri, 2002-04-12 at 16:42, Gene Grimm wrote: > I have simple question that I can't find documented. I need to add basic > spam/virus filters to body_checks using regular expressions where the > line begins with a tab character. Using "\t" or "\s" doesn't seem to > work. Can anyone suggest the best way to handle this? Are you using regexp or pcre maps ?? If you are using pcre style regular expressions then \t should work. If the map is a regexp style map, then you just use a tab instead of \t. Testing your regular expressions with postmap -q is a great way to find out any problems with the regexes. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mysql
On Tue, 2002-04-09 at 11:56, Fabrizio Tivano wrote: > MYD, .MYI and .frm files, in your newdbasesystem. > > > > I doubt this will work, because those binary files are not in a platform > > independend format and the format even varies between mysql versions. > > > > > Also there is al lot of ways and utilities in the mysq.com site on how-to > > > replicate and syncronize database. > it work... i've been tested importing a dbase from MySQL 3.23.49 > to 3.23.39 whitout loss any loss of data and error. Well, a while ago I helped someone rescue his mysql databases because he only had the .myd, .myi and .frm files and lost the mysql version that produced those files. Because I still had a 3.22.x mysqld running on a Linux/Alpha system, I could help him. Importing those files in a 3.23.x or Linux/i386 system would not work. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mysql
On Tue, 2002-04-09 at 10:28, Fabrizio Tivano wrote: > On Friday 08 February 2002 23:53, Erol wrote: > > I have a mysql server on FreeBsd OS and I want to > > transfer all the database to Debian Linux. > > > > Is it possible and How can I do that ? > for me the best way to do that is: > go in the mysql dbase directory (ex. /var/lib/mysq/your_dbase) > and brutally copy recursively, your_dbase dir that contains all your .MYD, > .MYI and .frm files, in your newdbasesystem. I doubt this will work, because those binary files are not in a platform independend format and the format even varies between mysql versions. > Also there is al lot of ways and utilities in the mysq.com site on how-to > replicate and syncronize database. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Spam filtering on the lists
On Fri, 2002-04-05 at 10:13, Jean-Marc V. Liotier wrote: > (Re: *SPAM* ()XX XX XX...) > > According to these headers contained in the message received from the > list, SpamAssassin has correctly tagged it as Spam. So why was it > forwarded anyway? Because it might not be spam, but only a false hit ?? Because blocking mail by the list could be considered some kind of sensorship ?? BTW, because you used multiple X's in 3 different words, your mail also scored 1.896 points for 'Uses words and phrases which indicate porn (3)'. This is not very much (and is compensated by using a In-Reply-To: header worth -5.029 points, so I did receive your message, but you get the idea :) I use spamassassin on my home and work mailservers (when I receive this message back from the mailinglist I know it has been checked 4 times by various spamassassins) and I really like it, but even spamassassin is not perfect enough to use it to really block mail. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bandwidth on Apache
On Thu, 2002-03-14 at 11:07, Michal Novotny wrote: > How can I slow down connection to/from one virtual host? mod_throttle -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: two ethernet without routing
On Wed, 2002-03-13 at 12:16, [EMAIL PROTECTED] wrote: > Well, there some kind of routing even: > > /etc/network/options: > ip_forward=no > > was already done. > I still can connet to my server through both IP's even the eth1 has no cable > connected. This behaviour is not routing. The server is not 'moving' a packet from one subnet to the other subnet. It is only just listening for every ip it knows on every interface. Use a firewall on the server to block unwanted connections (or better, use a firewall to only allow wanted connections) > From: "Stephane Bortzmeyer" <[EMAIL PROTECTED]> > > On Wed, Mar 13, 2002 at 11:27:31AM +0100, > > [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote > > a message of 18 lines which said: > > > > > How can I deactivate the routing option betwen cards? > > > > /etc/network/options: > > > > ip_forward=no > > > > Check with 'cat /proc/sys/net/ipv4/ip_forward'. > > > > > There have to be some kind of routing now because I can connect to my > apache > > > typing the two IPs even I've just one cable connected to eth0. > > > > Wrong analysis. As soon as the machine receives a packert with one of > > its IP addresses as the destination, it will handle it (even if it > > does not come from the "right" network card). -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Start Up Time
On Tue, 2002-02-19 at 22:43, Kupy Lopez wrote: > Hi. This is the first time I decided to write to a debian list, so that, please don't hate me by my bad english and my shilly questions. ;-) Next time you post, try not to post using html. And without those long lines. I'm viewing this on a 24" widescreen monitor and it doesn't fit on my screen :) > My problem is that, when I start my e-mail server (debian potato, with postfix, and qpopper), the pop3 queries are too long. In fact, programs like Outlook express try to download new emails but they can't because they get the tipical error telling that the server is not respoding within 60 seconds. If you answer to outlook that you want to wait for 60 seconds more, the server responds. I have this problem only when I rebbot the server. About 5 days later, the server responds normally within the first 60 seconds. I get a similar error in an other server (debian potato too) but this time, the server takes a long time to respond any tipe of service query until it has no more free RAM. I don't know why the servers respond better wen they have no free RAM. Could anybody help me? This sounds like you are having nameserver problems. I guess your popserver is trying to look up the hostname for the client that connects and this takes a while (2 minutes ??) before it times out. These lookups are probably done because you have an ALL:PARANOID line in your /etc/hosts.deny, which is a good thing, but you could remove this line and see if it solves the problem. If it does solve the problem, then you should investigate what goes wrong with the name server lookups and fix the real problem. -- Tot ziens, Bart-Jan
Re: Start Up Time
On Tue, 2002-02-19 at 22:43, Kupy Lopez wrote: > Hi. This is the first time I decided to write to a debian list, so that, please don't hate me by my bad english and my shilly questions. ;-) Next time you post, try not to post using html. And without those long lines. I'm viewing this on a 24" widescreen monitor and it doesn't fit on my screen :) > My problem is that, when I start my e-mail server (debian potato, with postfix, and qpopper), the pop3 queries are too long. In fact, programs like Outlook express try to download new emails but they can't because they get the tipical error telling that the server is not respoding within 60 seconds. If you answer to outlook that you want to wait for 60 seconds more, the server responds. I have this problem only when I rebbot the server. About 5 days later, the server responds normally within the first 60 seconds. I get a similar error in an other server (debian potato too) but this time, the server takes a long time to respond any tipe of service query until it has no more free RAM. I don't know why the servers respond better wen they have no free RAM. Could anybody help me? This sounds like you are having nameserver problems. I guess your popserver is trying to look up the hostname for the client that connects and this takes a while (2 minutes ??) before it times out. These lookups are probably done because you have an ALL:PARANOID line in your /etc/hosts.deny, which is a good thing, but you could remove this line and see if it solves the problem. If it does solve the problem, then you should investigate what goes wrong with the name server lookups and fix the real problem. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bandwidth control on FTP
On Wed, 2002-02-13 at 19:57, Michael Merritt wrote: > I've got some files on an anonymous FTP server that I'd like to open up for > download to users of some online forums. But I don't want to let them > completely suck my bandwidth dry. I'm running proftpd 1.2.4. > > I see in the documentation there are bandwidth limits like this: > RateReadBPS 81920 > RateReadFreeBytes 5120 > RateReadHardBPS on > > But that will limit each client to n bps, regardless if they are on 56k or > T3. I don't want to do this; instead, I want to allow say 1Mbps of bandwidth > for upload and let whoever has the big pipes get the fast downloads. Can't you use the kernel traffic shaper for this ?? -- Tot ziens, Bart-Jan
Re: Bandwidth control on FTP
On Wed, 2002-02-13 at 19:57, Michael Merritt wrote: > I've got some files on an anonymous FTP server that I'd like to open up for > download to users of some online forums. But I don't want to let them > completely suck my bandwidth dry. I'm running proftpd 1.2.4. > > I see in the documentation there are bandwidth limits like this: > RateReadBPS 81920 > RateReadFreeBytes 5120 > RateReadHardBPS on > > But that will limit each client to n bps, regardless if they are on 56k or > T3. I don't want to do this; instead, I want to allow say 1Mbps of bandwidth > for upload and let whoever has the big pipes get the fast downloads. Can't you use the kernel traffic shaper for this ?? -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: RAID Suggestion for webserver
On Sun, 2002-02-10 at 23:09, Jason Lim wrote: > > > Okay, as you said, with RAID10 and 4 40G HDs, usable space is 80Gs. > > > > > > On the other hand, with RAID5 and 3 40G HDs, usable space is also > 80Gs, > > > with 1 spare HD for rebuilding. > > > > > > The question becomes... which provides more performance and is more > > > reliable? > > > > RAID10 will give you the most performance. Not only do you have 4 disks > > working for you all the time, instead of 3 with RAID5, you (or more > > accurate: your CPU) also don't have to calculate the parity which is > > used by RAID5. > > The CPU won't be handling this... the 3ware RAID card (hardware) will > perform the parity calculations, so RAID 5 won't cause that type of > slowdown due to additional CPU utilization. Mmm, this is one of the rare IDE RAID cards that are true hardware RAID. But you still have only 3 disks doing the work. How about the number of disk-IO operations needed to perform a read or write operation?? > > Both will survive a 1 disk crash with no problems and both will appear > > as a RAID0 array when running in degraded mode. However, the reliability > > is different when a second disk fails. In RAID5 with spare you are out > > of luck when a second disk fails while the spare is rebuilding. > > With RAID10 and 1 failed disk, you only have the disk that is in the > > same stripe as the failed disk that is save to fail. > > I'm not sure if the raid card supports a stripe of two mirrors. This > > setup will survive a 2 disk failure. > > With the RAID5 with 3 disks and 1 spare, the only time the array would be > vunerable would be during reconstruction onto the spare disk. Once that is > done, the array will be fully restored, and could survive another failure > in any disk. > > As you mentioned, the RAID10 with 4 disks could also survive 2 failures, > however (again as you mentioned) the 2nd disk that failures cannot be part > of the same stripe as the original failed disk. > > Performance while the RAID5 array is degraded won't be too bad due to the > fact this is hardware RAID and not software RAID, and the hardware's > dedicated RAID chips will handle the computations. In degraded mode there are no computations to be made. It's just the same as RAID0. The card just reads whatever is on the two disks and only sorts the blocks in the right order. The same is with the RAID10 setup. Although this setup still has 3 disks, one of those 3 only has half of a stripe and can't be used in the array, so in this case the degraded RAID10 is also something like a RAID0. The hardware's chips will be used for rebuilding the array on the spare disk. How much this will impact the normal operation depends on how fast you want the rebuild to take place. And even with dedicated chips this will degrade normal operation because it needs to read all information on the 2 disks in order to calculate what to put on the 3rd disk. > It seems RAID5 would be a safer solution as long as another failure does > not occur during the reconstruction onto the spare. Hum... how long would > the reconstruction take for a 40G hd? I'm guessing 30-40 minutes? Would > that be about right? I'm not sure how fast today's disks really are, but remember that the rebuild time also depends on how much normal work the array has to do (I asume the card has a setting that allows you to tune how the card divides the rebuilding work versus the normal work). -- Tot ziens, Bart-Jan
Re: RAID Suggestion for webserver
On Sun, 2002-02-10 at 22:38, Jason Lim wrote: > Okay, as you said, with RAID10 and 4 40G HDs, usable space is 80Gs. > > On the other hand, with RAID5 and 3 40G HDs, usable space is also 80Gs, > with 1 spare HD for rebuilding. > > The question becomes... which provides more performance and is more > reliable? RAID10 will give you the most performance. Not only do you have 4 disks working for you all the time, instead of 3 with RAID5, you (or more accurate: your CPU) also don't have to calculate the parity which is used by RAID5. Both will survive a 1 disk crash with no problems and both will appear as a RAID0 array when running in degraded mode. However, the reliability is different when a second disk fails. In RAID5 with spare you are out of luck when a second disk fails while the spare is rebuilding. With RAID10 and 1 failed disk, you only have the disk that is in the same stripe as the failed disk that is save to fail. I'm not sure if the raid card supports a stripe of two mirrors. This setup will survive a 2 disk failure. -- Tot ziens, Bart-Jan
Re: RAID Suggestion for webserver
On Sun, 2002-02-10 at 23:09, Jason Lim wrote: > > > Okay, as you said, with RAID10 and 4 40G HDs, usable space is 80Gs. > > > > > > On the other hand, with RAID5 and 3 40G HDs, usable space is also > 80Gs, > > > with 1 spare HD for rebuilding. > > > > > > The question becomes... which provides more performance and is more > > > reliable? > > > > RAID10 will give you the most performance. Not only do you have 4 disks > > working for you all the time, instead of 3 with RAID5, you (or more > > accurate: your CPU) also don't have to calculate the parity which is > > used by RAID5. > > The CPU won't be handling this... the 3ware RAID card (hardware) will > perform the parity calculations, so RAID 5 won't cause that type of > slowdown due to additional CPU utilization. Mmm, this is one of the rare IDE RAID cards that are true hardware RAID. But you still have only 3 disks doing the work. How about the number of disk-IO operations needed to perform a read or write operation?? > > Both will survive a 1 disk crash with no problems and both will appear > > as a RAID0 array when running in degraded mode. However, the reliability > > is different when a second disk fails. In RAID5 with spare you are out > > of luck when a second disk fails while the spare is rebuilding. > > With RAID10 and 1 failed disk, you only have the disk that is in the > > same stripe as the failed disk that is save to fail. > > I'm not sure if the raid card supports a stripe of two mirrors. This > > setup will survive a 2 disk failure. > > With the RAID5 with 3 disks and 1 spare, the only time the array would be > vunerable would be during reconstruction onto the spare disk. Once that is > done, the array will be fully restored, and could survive another failure > in any disk. > > As you mentioned, the RAID10 with 4 disks could also survive 2 failures, > however (again as you mentioned) the 2nd disk that failures cannot be part > of the same stripe as the original failed disk. > > Performance while the RAID5 array is degraded won't be too bad due to the > fact this is hardware RAID and not software RAID, and the hardware's > dedicated RAID chips will handle the computations. In degraded mode there are no computations to be made. It's just the same as RAID0. The card just reads whatever is on the two disks and only sorts the blocks in the right order. The same is with the RAID10 setup. Although this setup still has 3 disks, one of those 3 only has half of a stripe and can't be used in the array, so in this case the degraded RAID10 is also something like a RAID0. The hardware's chips will be used for rebuilding the array on the spare disk. How much this will impact the normal operation depends on how fast you want the rebuild to take place. And even with dedicated chips this will degrade normal operation because it needs to read all information on the 2 disks in order to calculate what to put on the 3rd disk. > It seems RAID5 would be a safer solution as long as another failure does > not occur during the reconstruction onto the spare. Hum... how long would > the reconstruction take for a 40G hd? I'm guessing 30-40 minutes? Would > that be about right? I'm not sure how fast today's disks really are, but remember that the rebuild time also depends on how much normal work the array has to do (I asume the card has a setting that allows you to tune how the card divides the rebuilding work versus the normal work). -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: RAID Suggestion for webserver
On Sun, 2002-02-10 at 22:38, Jason Lim wrote: > Okay, as you said, with RAID10 and 4 40G HDs, usable space is 80Gs. > > On the other hand, with RAID5 and 3 40G HDs, usable space is also 80Gs, > with 1 spare HD for rebuilding. > > The question becomes... which provides more performance and is more > reliable? RAID10 will give you the most performance. Not only do you have 4 disks working for you all the time, instead of 3 with RAID5, you (or more accurate: your CPU) also don't have to calculate the parity which is used by RAID5. Both will survive a 1 disk crash with no problems and both will appear as a RAID0 array when running in degraded mode. However, the reliability is different when a second disk fails. In RAID5 with spare you are out of luck when a second disk fails while the spare is rebuilding. With RAID10 and 1 failed disk, you only have the disk that is in the same stripe as the failed disk that is save to fail. I'm not sure if the raid card supports a stripe of two mirrors. This setup will survive a 2 disk failure. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian Archive Mgmt
On Wed, 2002-01-30 at 07:21, [EMAIL PROTECTED] wrote: >Im almost sure there is a script/program somewhere there that creates >Packages.gz and Release files for a given set of .debs in order to use >them in apt/sources.list > >My current script does > >dpkg --info |grep relevant fields >calculate remaining fields (filename,size, md5 etc.)>> Packages > >And somewhat does create a 'normal' Packages file, with some fields swapped >in a different order than in a debian.org packages, but apt-get update >does not like my Packages file, is states that it can't parse it and quits. Why not use dpkg-scanpackages ?? apt-get install dpkg-dev to get dpkg-scanpackages. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt
On Sat, 13 Oct 2001, Kevin wrote: > is there a way to lock a package so that apt/dpkg wont update it? > > i use a bofh'd bash, but it keeps getting overwritten by new bash > packages. i suppose i could chattr +i it but im hoping theres a more > elegant solution. If your bash version is also a dpkg package, then putting it on hold is the way to go. If it's just a single binary, then dpkg-divert can help. -- Tot ziens, Bart-Jan
Re: apt
On Sat, 13 Oct 2001, Kevin wrote: > is there a way to lock a package so that apt/dpkg wont update it? > > i use a bofh'd bash, but it keeps getting overwritten by new bash > packages. i suppose i could chattr +i it but im hoping theres a more > elegant solution. If your bash version is also a dpkg package, then putting it on hold is the way to go. If it's just a single binary, then dpkg-divert can help. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: BIND
On Wed, 3 Oct 2001, Craig wrote: > What are the standard entries in a zone file suppose > to be ? ie the serial no and TTL and refresh that ISP's > use ? RFC1537 and RFC1912 discuss common DNS errors and solutions, including valid values for serial numbers and sane values for the other SOA fields. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
RE: Sendmail or DNS Problem?
On Mon, 27 Aug 2001, Chad Morgan wrote: > Here is my sendmail.mc file. > > > divert(-1) > # > # Copyright (c) 1983 Eric P. Allman > # Copyright (c) 1988, 1993 > # The Regents of the University of California. All rights reserved. > # > > # > # This file is used to configure sendmail for use with Debian systems. > # > > divert(0) > VERSIONID(`@(#)sendmail.mc8.9.3-21 (Debian) 2309') > OSTYPE(debian)dnl > LOCAL_CONFIG > define(`SMART_HOST', `N')dnl There you have the problem. SMART_HOST should be defined like this: define(`SMART_HOST', `mailer:hostname') You told sendmail to send all it's mail to a host called N (sendmail did convert it to a lowercase n however. > FEATURE(always_add_domain)dnl > Cwns.enhancetheweb.com > FEATURE(use_cw_file)dnl > FEATURE(use_ct_file)dnl > FEATURE(nouucp)dnl > MAILER_DEFINITIONS > MAILER(local)dnl > MAILER(smtp)dnl > > LOCAL_CONFIG > MASQUERADE_AS(ns.enhancetheweb.com)dnl > ## Custom configurations below (will be preserved) > FEATURE(promiscuous_relay) > ###EOF -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: e2fsck error
On Fri, 3 Aug 2001, Gene Grimm wrote: > Bart-Jan Vrielink wrote: > > On Thu, 2 Aug 2001, Gene Grimm wrote: > > > > A short while ago, my web server went offline and is now reporting a > > > > superblock error on /dev/hdc1 (/var) and suggests using e2fsck -b > > > > 8193 option. I do not know enough about the ext2 system to know > > > > what super block numbers to try. Can anyone suggest the correct > > > > numbers or how to find out? This is a personal web server plus I use > > > > this to test configurations. > > > > > > > To follow up, I have tried several numbers from 8193 incrementing by > > > 8192, and they all report something like invalid magic number in super > > > block. > > > > > Please read the e2fsck manpage. The exact position of the backup > > superblock depends on the blocksize of your filesystem. You can find the > > blocksize by running dumpe2fs -h /dev/hdc1| grep "Block size" > > dumpe2fs -h /dev/hdc1 reports: > Bad magic number in super-block while trying to open /dev/hdc1. > Couldn't find valid filesystem superblock. > > Since I don't know much about the ext2 file system, I am afraid to use the > debugfs utility without specific guidence. Sorry, my mistake (dumpe2fs uses the first superblock to get it's information...) If the filesystem was initialized with mke2fs using the default settings it's very likely that it's using a 4k blocksize, so the backup superblock should be at 32768. If you're afraid to use tools you're not familiar with on that filesystem, why not copy that filesystem to a file using dd and test the tools on that file first ?? -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need advise: productive HDD is down
On Tue, 31 Jul 2001, Russell Coker wrote: > On Mon, 30 Jul 2001 16:16, Dmitry Litovchenko wrote: > > fsck (e2fsck) refuses to fix anything telling different things on > > different partitions, "Cannot mount FAT due to some blah blah blah" or > > "superblock is damaged try to run e2fsck -b 8193 blah blah blah" which > > is also failed (I mean e2fsck -b 8193). > > debugfs is one program that may be able to help. Also trying the -b8193 > option may help too... Some time ago I managed to destroy my /home (ran mkswap on it). e2fsck also failed on it and suggested to do a -b 8193 which also failed. After reading the manpage I found out that since a few years the position of the backup superblocks depend on the blocksize. For filesystems with 1k blocksizes, a backup superblock can be found at block 8193; for filesystems with 2k blocksizes, at block 16384; and for 4k blocksizes, at block 32768. Maybe using -b 32768 will help ?? -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ATA Speed
On Tue, 3 Jul 2001, R K wrote: > Does the following mean that Linux is only using my ide bus at ata33 > speeds? Or more accurately not using the full ata100 mode? > > ide: Assuming 33MHz system bus speed for PIO modes; override with > idebus=xx > > I've seen nothing from dmesg to indicate that it's doing > otherwise. Does it configure it as 33 and then still use it to it's > full potential or does it impose restrictions on itself? Even if this > doesn't have anything to do with it, how would I verify that Linux is > using the hardware to its full potential? Did you note that it said 33MHz and not 33 MB/s ?? And did you note that it says PIO mode, while ata100 is a DMA mode ?? Entering that message into Google got me the following url: http://list.cobalt.com/pipermail/cobalt-users/2001-May/042555.html which quotes another message (I am too lazy to find the original) which clearly explains what this message means. -- Tot ziens, Bart-Jan
Re: ATA Speed
On Tue, 3 Jul 2001, R K wrote: > Does the following mean that Linux is only using my ide bus at ata33 > speeds? Or more accurately not using the full ata100 mode? > > ide: Assuming 33MHz system bus speed for PIO modes; override with > idebus=xx > > I've seen nothing from dmesg to indicate that it's doing > otherwise. Does it configure it as 33 and then still use it to it's > full potential or does it impose restrictions on itself? Even if this > doesn't have anything to do with it, how would I verify that Linux is > using the hardware to its full potential? Did you note that it said 33MHz and not 33 MB/s ?? And did you note that it says PIO mode, while ata100 is a DMA mode ?? Entering that message into Google got me the following url: http://list.cobalt.com/pipermail/cobalt-users/2001-May/042555.html which quotes another message (I am too lazy to find the original) which clearly explains what this message means. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Check NIC for speed and promisc mode
On Tue, 19 Jun 2001, Jason Lim wrote: > How can I check the "real" connection speed of a NIC? These are realtek > 8129/8139 network cards. The leds behind the NIC aren't exactly > informative. I was hoping there was some way to do this directly in linux > through a hardware call. I checked /proc and i couldn't find anything > pertaining to this. mii-tool (Package: net-tools) or mii-diag (Package mii-diag). > On a related note, how can I check if the card is in "promisc" mode? Some > software seems to leave it in promisc mode... so once I find out if it IS > in promisc mode, how can i switch it back? ifconfig eth0 | grep PROMISC ifconfig eth0 -promisc -- Tot ziens, Bart-Jan
Re: Check NIC for speed and promisc mode
On Tue, 19 Jun 2001, Jason Lim wrote: > How can I check the "real" connection speed of a NIC? These are realtek > 8129/8139 network cards. The leds behind the NIC aren't exactly > informative. I was hoping there was some way to do this directly in linux > through a hardware call. I checked /proc and i couldn't find anything > pertaining to this. mii-tool (Package: net-tools) or mii-diag (Package mii-diag). > On a related note, how can I check if the card is in "promisc" mode? Some > software seems to leave it in promisc mode... so once I find out if it IS > in promisc mode, how can i switch it back? ifconfig eth0 | grep PROMISC ifconfig eth0 -promisc -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: CGI Errors
On Tue, 12 Jun 2001, Ritesh Goel wrote: > I have a script which runs fine from the UNIX shell. As soon as I try > going thru the browser, I get an error which says - "Premature end of > Script Header". What should i do or what can i do ?? It looks like the script does not produce any valid HTTP headers. As a minimum it should produce a line 'Content-Type: text/html' (or any other content-type) and a blank line, before the html begins. -- Tot ziens, Bart-Jan
Re: CGI Errors
On Tue, 12 Jun 2001, Ritesh Goel wrote: > I have a script which runs fine from the UNIX shell. As soon as I try > going thru the browser, I get an error which says - "Premature end of > Script Header". What should i do or what can i do ?? It looks like the script does not produce any valid HTTP headers. As a minimum it should produce a line 'Content-Type: text/html' (or any other content-type) and a blank line, before the html begins. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc and cpp errors
On Thu, 10 May 2001, Nathan Ridge wrote: > has anyone seen the following errors before? : > > gcc: Internal compiler error: program cc1 got fatal signal 4 > gcc: Internal compiler error: program cc1 got fatal signal 11 > cpp0: output pipe has closed > > I am trying to compile kernel-source-2.2.17, seems to be very random, I can > simply try again, and it may go further, until it get stucks on a certain .c > file, then after a reboot, will go a little further. Have tried stable and > unstable packages. Any help appreciated. See http://www.bitwizard.nl/sig11/ -- Tot ziens, Bart-Jan
Re: gcc and cpp errors
On Thu, 10 May 2001, Nathan Ridge wrote: > has anyone seen the following errors before? : > > gcc: Internal compiler error: program cc1 got fatal signal 4 > gcc: Internal compiler error: program cc1 got fatal signal 11 > cpp0: output pipe has closed > > I am trying to compile kernel-source-2.2.17, seems to be very random, I can > simply try again, and it may go further, until it get stucks on a certain .c > file, then after a reboot, will go a little further. Have tried stable and > unstable packages. Any help appreciated. See http://www.bitwizard.nl/sig11/ -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: htaccess failure
On Mon, 19 Feb 2001, Martin WHEELER wrote: > To my knowledge, protecting a directory using the .htaccess file method > has always worked without a hitch for me on my local machine (Debian > 2.2r2 + proposed-updates) -- but has "suddenly" stopped functioning. > > [EMAIL PROTECTED]:~$ apache -v > Server version: Apache/1.3.9 (Unix) Debian/GNU > Server built: Jan 26 2001 00:10:13 > > is what I'm running (stable hasn't progressed to 1.3.12 yet); the > access.conf file contains: > > > Options Includes Indexes FollowSymLinks > AllowOverride None The line above means "ignore any .htaccess files in this directory". I don't think you'll want that :) I think you'll want that line to read AllowOverride AuthConfig > order allow,deny > allow from all > > > and: > > # Do not allow retrieval of the override files, a standard security > measure. > > order allow,deny > deny from all > > > srm.conf contains: > > # AccessFileName: The name of the file to look for in each directory > # for access control information. > AccessFileName .htaccess > > but no way will Apache stop and ask for authentication when any user > goes into any directory containing a valid .htaccess file pointing to a > valid htusers/.htpasswd data file. -- Tot ziens, Bart-Jan
Re: htaccess failure
On Mon, 19 Feb 2001, Martin WHEELER wrote: > To my knowledge, protecting a directory using the .htaccess file method > has always worked without a hitch for me on my local machine (Debian > 2.2r2 + proposed-updates) -- but has "suddenly" stopped functioning. > > mwheeler@startext:~$ apache -v > Server version: Apache/1.3.9 (Unix) Debian/GNU > Server built: Jan 26 2001 00:10:13 > > is what I'm running (stable hasn't progressed to 1.3.12 yet); the > access.conf file contains: > > > Options Includes Indexes FollowSymLinks > AllowOverride None The line above means "ignore any .htaccess files in this directory". I don't think you'll want that :) I think you'll want that line to read AllowOverride AuthConfig > order allow,deny > allow from all > > > and: > > # Do not allow retrieval of the override files, a standard security > measure. > > order allow,deny > deny from all > > > srm.conf contains: > > # AccessFileName: The name of the file to look for in each directory > # for access control information. > AccessFileName .htaccess > > but no way will Apache stop and ask for authentication when any user > goes into any directory containing a valid .htaccess file pointing to a > valid htusers/.htpasswd data file. -- Tot ziens, Bart-Jan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
