| |
| |
| |
|
Page: 1 2 3
Comments:
<0> SmooveB thanks
<1> which command add a user to a specific group?
<2> take a guess
<1> what?
<2> c'mon. it'll be fun
<1> ah , thanks ;)
<3> heh
<3> daowee, you rock
<4> I am installing openbsd as a dedicated transparent firewall on a 9GB HD, can someone recommend a good partitioning scheme? I would like to make /var as large as possible but I do not want to make the others too small and would like to find out how much space is needed.
<5> / 512M, /tmp 256M, /usr 3G, /home 1G, /var rest
<4> swap?
<5> how much ram?
<4> 1GB.
<5> 1.05 G
<5> it's desirable to have swap slightly larger than ram, in order to get crashdumps should the system become unstable
<5> though it's unusual to have a smallish drive in a machine with so much ram..
<4> makes sense
<4> i was hoping to ahve more than 4gb in /var but i guess i could just archive the logs every night to an offsite server
<5> even if you got a crashdump, you could only store 2 in that relatively big /var, for instance
<5> you can trim down /home first, and /usr too.. depends how much space you want available for packages, or if you intend to rebuild a snapshot yourself at some point
<5> I tend to leave enough room in /usr to do a rebuild
<5> without a lot of effort you can fit everything into 512M.. but you sacrifice flexibility and generally just cause more work in the future
<4> figz, heh, the server is a sun netra with 2x18GB SCSI in it and 1GB RAM, but I replaced the 18GB with 9GB since I had so many in stock
<4> I know the HD's are small but I was hoping it'd be all I need since it willbe used as a dedicated firewall
<4> perhaps i should put the 18gb's back in....
<5> put another 9GB in as your /var
<6> anyong running snort + prelude
<7> mmm... new snap on zaurus gives xf86OpenConsole: KDENABIO failed (Operation not permitted) on startx. i see there's no recent x snap packages. my latest prior was from about a month ago. i suppose this could be the reason. anyone seeing this?
<4> Figz, would putting in a second 9GB for /var be better than putting in one 18GB? I figure yes for performance reasons since /var is on a seperate physical disk
<2> depends.. the 18GB might be a newer generation.. and you probably have zero parallel load on the fws disk system anyway
<2> two disks means twice the failure rate
<5> but you mentioned that you had an excess of 9GB disks...
<8> does pkgsrc work on OpenBSD?
<9> when i run /usr/local/bin/mysqld_safe it promts that the server has started but then nothing happens, even if i press ^C or ^Z, same thing happens on boot but i am capable of ssh into the box
<9> what might cause the error?
<9> i have followed this guide: http://www.openbsdsupport.org/mysql.htm
<9> no sorry
<9> this one: http://www.openbsdsupport.org/oamp.pdf
<10> check if the server is running (ps), check its log file /var/db/mysql/*.err
<10> if everything is fine and the only issue is that the mysqld_safe script is not detaching, run it with & ...
<9> the server is running
<9> ok ill check
<10> i think that pdf is simply missing the trailing & on the rc.local line, it uses it in the subsequent example for a command line startup
<11> timo: mysqld_save won't background use ...../mysqld_save &
<9> i'll ssh to the machine, kill mysqld_safe and try to runt with &
<11> yes
<9> -bash-3.1# mysqld_safe &
<9> [1] 18032
<9> -bash-3.1# Starting mysqld daemon with databases from /var/mysql
<9> then nothing happens
<11> what do you expect to happen?
<10> that's fine, nothing more is expected to happen :)
<9> but .. when booting i would like it to keep booting :)
<9> it freezes
<10> so add that & to the line in rc.local, too
<9> but the & parameter didnt change anything
<9> it still freezes
<10> no, you got another bash prompt, didn't you?
<11> hit return
<9> well
<9> it continues when i hit return..
<9> but it is supposed to be like that on boot?
<10> the boot sequence will continue, try it
<10> you won't have to press return or such
<9> just add &
<11> yes
<9> lemme check, be right back
<10> (at the end of the line, right-most side, not before other parameters)
<9> ?
<9> if [ X"${mysql}" == X"YES" -a -x /usr/local/bin/mysqld_safe & ]; and /usr/local/bin/mysqld_safe & --user=_mysql \
<9> correct?
<11> no
<11> /usr/local/bin/mysqld_safe --user=_mysql &
<11> and
<11> -x /usr/local/bin/mysqld_safe & remove it here
<10> post the second line, too, the \ implies the line continues...
<9> m_athias: and & remove it there?
<11> mysqld_safe & ]; and
<9> if [ X"${mysql}" == X"YES" -a -x /usr/local/bin/mysqld_safe & ]; (remove what)
<11> remove the & there
<9> aha
<9> but the line begins with if [
<10> if [ ... -a -x file ]; means "if ... and file exists"
<10> that's just checking if the binary exists, it doesn't start it yet, hence the & there makes no sense
<9> if [ X"${mysql}" == X"YES" -a -x /usr/local/bin/mysqld_safe;
<10> look at http://postfixwiki.org/index.php?title=OpenBSD_PostfixAdmin_Guide
<10> it seems you need a careful copy/paste ;)
<9> but i would be glad if you could help me with just these two lines so i can boot the machine without any problems ;)
<10> so paste the entire section of rc.local related to mysql
<9> i'll look into it
<9> thanks
<10> ...
<9> sorry for all the stupid questions with quick fixes, but i have two weeks vaccation where i'm supposed to help another company to set up a production e-mail server with webmail access
<10> it's not a stupid question, but the ***umption that we can give you the precisely right instructions _before_ seeing the precise section you have now is flawed
<9> i see
<10> you need to add precisely one & character in the precisely right place :)
<9> hehe
<10> not just sprinkle it in "somewhere there" :)
<9> i'll read it all, maybe postfix would be an application i need on this server
<10> it's not about postfix, it has an rc.local example for mysql
<9> okey
<9> do you have any suggestions on easy to use e-mail server applications that i can use in conjuction with the webserver for webmail access?
<11> timo: postfix is 'easy' to start (compared to sendmail)
<11> but depends what you need
<9> i don't need any advanced features, approx <200 users who would want to use e-mail in their clients and also have access to webmail; my isp has blocked port 25 so i have to relay the outgoing e-mail in some way
<9> probably i have to use my isp's smtp-server
<9> so i have to setup my local smtp to forward all e-mails to that one (my isp)
<11> it's easy with postfix
<9> i'll look into it; i guess i have to disable sendmail but i guess it's all explained in the postfix documentation
<9> /usr/local/bin/mysqladmin -u root -h myserver.mydomain.com p***word 'new-p***word' - i don't know what to fill in at myserver.mydomain.com, because i don't have a domain yet
<9> error: 'Host '-whatever i type-' is not allowed to connect to this MySQL server'
<11> try mysql -u root
<11> you should be able to connect without p***word the first time
<11> than you can set your p***word from there
<9> i successfullt set my p***word the first time, then when i executed the othe command it failed
<11> can you connect with 'mysql -u root -p'
<11> ?
<9> yes
<11> so whats the problem?
<9> i can't do the other command: /usr/local/bin/mysqladmin -u root -h myserver.mydomain.com p***word 'new-p***word'
<9> that was outlined as a requirement
<11> wince you changed your root p***word already I think you don't need it
<11> you can try
<11> mysqladmin -u root -p -h localhost p***word 'the new p***word you want to set'
<11> It should ask you for your old p***word, and set the new one
<12> anyone know why pf would give an error for this line, p*** out on $ext_if proto icmp from $ext_if to any icmp-type 8 keep state -> on pfctl -f /etc/pf.conf -> /etc/pf.conf:160: proto icmp doesn't match address family inet6
<9> i have this strange, perhaps stupid, problem; have an openbsd gateway and all of all sudden i couldn't ping my isp's dns so i restarted the machine and it worked fine again - my main problem is that i'm unable to ping my isp's dns inside my internal network but i can communicate well (ssh) to my gateway and ping from there
<11> changed anything on the pf ruleset?
<9> no
<9> all of a sudden the ssh connections from one client aborts (software caused connection abort)
<11> aggressive timeouts?
<9> maybe the switch is over heated, or some hardware error on the gateway
<9> what do you mean by aggressive timeouts?
<11> you loose connection to the gateway or to some external host?
<9> first thing that happens was that i lost (ssh) connection to my gateway but was able to establish it again, second thing was that i was unable to ping my isp's dns from my gateway so i rebooted it and then i could ping the dns again (strange..), third thing is that i am completely able to ping my isp's dns from the gateway but not from the clients inside my lan + the ssh clients to my gateway causes connection aborts
<11> if you didn't change anything I would check hardware
<11> atach a screen if you can and watch logs/dmesg/etc for strange things
<9> dmesg looks fine
<9> what about the other things
<11> check it after/while you have problems
<9> ill try to replace the switch right away
<9> the one i used has been placed on a very hot computer
<9> has been using
<9> can that cause this kind of problems
<9> problem remains
<9> how can i attach a screen and check for logs about strange things?
<13> can anyone tell me if .. when using altq and queue's, does it work if I ***ign a queue to in incoming rule with keep state so the traffic going back out on that connection/port get queued?
<11> you plug in a normal screen+keyboard, login and use tail -f ?
<9> m_athias: tail on what file
<11> I would start on /var/log/messages
Return to
#openbsd
or
Go to some related
logs:
freebsd snd_ich i915
klite dmz
#nhl
#worldcup
#3dsmax
#sql
gt6600 s-video
hefmail
banned by araditracker
td/tool
|
|