1257380812 Q * yarihm Quit: This computer has gone to sleep 1257383047 Q * dowdle Remote host closed the connection 1257384575 Q * scientes__ Ping timeout: 480 seconds 1257386077 Q * SubZero 1257389759 J * uva bno@118-168-239-250.dynamic.hinet.net 1257391891 Q * yepsen Ping timeout: 480 seconds 1257391926 J * yepsen ~yepsen@pdbn-5d825b32.pool.mediaWays.net 1257392755 Q * balbir_ Ping timeout: 480 seconds 1257393556 J * saulus_ ~saulus@c152253.adsl.hansenet.de 1257393916 Q * SauLus Read error: Connection reset by peer 1257393921 N * saulus_ SauLus 1257394864 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1257399205 M * Bertl off to bed now .. have a good one everyone! 1257399211 N * Bertl Bertl_zZ 1257400580 M * micah q/win 5 1257400582 M * micah q/win 53~ 1257400585 M * micah arg 1257405300 J * yarihm ~yarihm@office-zrh.youngsolutions.ch 1257405428 M * blues re 1257405849 M * blues Bertl_zZ: on that particular system its: 1257405853 M * blues # cat /proc/virtual/info 1257405853 M * blues VCIVersion: 0002:0304 1257405853 M * blues VCISyscall: 236 1257405853 M * blues VCIKernel: 13000ff1 1257405883 M * blues Bertl_zZ: kernel 2.6.27.12 1257405897 M * blues Bertl_zZ: but checked on never systems - no change 1257405937 M * blues Bertl_zZ: I think it should be called "feature request", not bug report :) 1257406275 J * sharkjaw ~gab@90.149.121.45 1257406843 Q * derjohn_mob Ping timeout: 480 seconds 1257407886 J * derjohn_mob ~aj@213.238.45.2 1257408000 J * AlexanderS ~Alexander@78.52.224.72 1257408106 M * AlexanderS hi 1257408476 M * AlexanderS hey, anybody knows how it is possible to have different vservers on the same host routing via different interfaces to different subnets? 1257408680 J * davidkarban ~david@199.123.broadband11.iol.cz 1257408836 Q * derjohn_mob Ping timeout: 480 seconds 1257408891 J * hijacker ~hijacker@213.91.163.5 1257409809 Q * _Shiva_ Quit: Reconnecting 1257409810 J * _Shiva_ shiva@whatcha.looking.at 1257409945 Q * _Shiva_ 1257409983 J * _Shiva_ shiva@whatcha.looking.at 1257410208 J * friendly ~friendly@ppp118-209-136-134.lns20.mel6.internode.on.net 1257410474 J * derjohn_mob ~aj@213.238.45.2 1257411910 Q * infowolfe Ping timeout: 480 seconds 1257413915 Q * sharkjaw Remote host closed the connection 1257413936 J * sharkjaw ~gab@90.149.121.45 1257414165 J * infowolfe ~infowolfe@97-120-149-152.ptld.qwest.net 1257414213 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1257414796 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1257415645 M * ghislainocfs2 vserver network is not different than normal routing it is just isolation all the networking is done on the host so if you can do it on a normal linux then you can do it on vserver :) 1257416080 J * alex__ ~Alexander@g225075212.adsl.alicedsl.de 1257416413 J * scientes ~scientes@174-21-153-224.tukw.qwest.net 1257416487 Q * AlexanderS Ping timeout: 480 seconds 1257416943 Q * friendly Quit: Leaving. 1257417581 Q * _Shiva_ Quit: Operator halted - Coffee not found 1257418793 J * _Shiva_ shiva@whatcha.looking.at 1257418919 Q * _Shiva_ 1257418921 J * _Shiva_ shiva@whatcha.looking.at 1257421557 P * alex__ 1257422421 J * doener_ ~doener@i59F56212.versanet.de 1257422521 Q * doener Ping timeout: 480 seconds 1257422724 Q * ex Quit: leaving 1257422732 J * ex ex@valis.net.pl 1257423145 J * SubZero ~SubZero@chello089076140236.chello.pl 1257425044 J * blues_ ~blues@actu185.neoplus.adsl.tpnet.pl 1257425162 Q * blues Ping timeout: 480 seconds 1257425895 Q * blues_ Quit: Reconnecting 1257425896 J * blues ~blues@actu185.neoplus.adsl.tpnet.pl 1257425931 J * thierryp ~thierry@zanzibar.inria.fr 1257426837 J * hparker ~hparker@208.4.188.201 1257428439 J * ryker jalberts@199.117.46.145 1257428468 M * ryker hi. Anyone have any luck running Cherokee web server in a guest? 1257428490 M * ryker Actually, the server is running ok, but I can't get cherokee-admin to work. 1257428598 Q * SubZero Read error: Connection reset by peer 1257428614 N * Bertl_zZ Bertl 1257428619 M * Bertl morning folks! 1257428622 J * SubZero ~SubZero@chello089076140236.chello.pl 1257428633 M * ryker good morning 1257428655 M * Bertl what's the problem with the admin? 1257428688 M * ryker it launches ok, but when I try to connect, it hangs 1257428692 M * ryker i used -x for debug 1257428704 M * ryker and I see the connection request 1257428712 M * ryker the default port is 9090 1257428720 M * ryker but the request says port 4000 1257428735 M * ryker Server 0.99.17 running.. PID=15470 Port=4000 1257428753 M * ryker i'm assuming that's just source port, so no biggie 1257428759 M * Bertl blues: did you set the IGNEG_NICE cflag? 1257428768 M * ryker no 1257428772 M * Bertl blues: http://linux-vserver.org/Capabilities_and_Flags 1257428772 M * ryker what does that do? 1257428804 M * ryker ignore priority raise? 1257428808 M * ryker what does that mean? 1257428831 M * Bertl it's for blues, it seems he requested a feature already there for some time :) 1257428845 M * ryker lol 1257428846 M * ryker sry 1257428877 M * ryker sry, it's early here. i ignored the blues part. ;) 1257428882 M * Bertl what does tcpdump -vvnei show on the host when you connect? 1257428893 M * ryker let me check 1257428907 M * Bertl (you might want to limit that to potential candidates) 1257428949 M * ryker bah 1257428963 M * ryker i just remembered, this is port 9090 and i need to open iptables for that guest on that port 1257428971 M * Bertl :) 1257428978 M * ryker sry for wasting your time 1257428990 M * ryker thx for your help 1257428993 M * Bertl no problem, glad it could be resolved and is not Linux-VServer related 1257429008 M * ryker no, just dumb*** related ;) 1257429422 Q * sharkjaw Quit: Leaving 1257429627 M * blues Bertl: thanks - you are great :D 1257429638 M * blues Bertl: exactly 1257429648 M * raceme ryker: don't be so rude it was just a keyboard/chair interface problem :) 1257429702 M * blues Bertl: shouldn't it be default? 1257429738 M * ryker raceme: :) 1257429764 M * Bertl blues: in general, we prefer to avoid ignoring things silently 1257429797 M * Bertl blues: usually this can be fixed without IGNEG_NICE, but adjusting the priorities/niceness accordingly 1257429809 M * ryker Bertl: actually, i'm still having the same problem, even after opening the port. I put the tidy'd up tcpdump output in pastebin at: http://pastebin.com/m52ba612d. Any chance you could take a look? It's all jiberish to me. 1257429810 M * Bertl (either inside the guest or for the guest on the host) 1257430021 M * Bertl ryker: you sure you used my suggested options? 1257430056 M * ryker I used: tcpdump -vvnei eth0 1257430067 M * Bertl anyway, looks like a 'normal' chat to me, did you try with telnet to the port? 1257430076 M * ryker and a couple grep statements to get rid of other fluff, like ssh 1257430116 M * ryker let me try again 1257430200 M * ryker Bertl: I had a grep in there for the IP I was coming from, is that ok? 1257430213 M * Bertl what I see is a chat between 199.117.46.144 and 207.56.64.194, noting that those are completely different IPs, does your server allow this kind of remote administration? 1257430283 M * ryker Yes, the 199.117.46.144 is the server, and 207.56.64.194 is our office IP. We all all ports between these 2 IP's. and I ran cherokee-admin to bind to all ports 1257430295 M * ryker I ran cherokee-admin -b -x 1257430319 M * ryker i even tried using a browser localhost and got the same result 1257430350 M * ryker this very well may be a cherokee-admin problem, but I've never used it before on a guest 1257430369 M * Bertl any messages on the server's syslog (or cherokee specific log)? 1257430434 M * ryker no, none at all 1257430446 M * ryker just some in the cherokee syslog for successfull connects on port 80 1257430457 M * ryker but isn't cherokee-admin completely seperate anyway? 1257430466 M * Bertl no idea, never used it 1257430502 M * ryker I think I will temp install it on the host 1257430505 M * ryker and see if it works 1257430507 M * Bertl well, I'd try to narrow it down via tcpdump based on the port 1257430559 M * Bertl it might be, that it does try to contact 'something' via localhost, which might be misconfigured, or your guest might have a single IP with single IP special casing enabled ... 1257430570 Q * infowolfe Ping timeout: 480 seconds 1257430604 M * Bertl I'd also try to enable debugging for the components involved, often the service itself knows exactly what goes wrong but doesn't say without being told to do so 1257430615 Q * BenG Quit: I Leave 1257430640 M * ryker afaik, the only debug option for cherokee-admin is passing the -x parm, which is how I started it. 1257430645 M * ryker it does show the initial connect 1257430648 M * ryker but that's about it. 1257430674 M * ryker i'm installing on the vserver host real quick just to see if cherokee-admin works 1257430775 M * ryker Bertl: works great on the host 1257430802 M * ryker Bertl: i'll do a tcpdump on the host and compare 1257431453 J * elesouef ~elesouef@212.51.177.50 1257431502 M * ryker Bertl: I don't know if this will help you help me at all, but here is a tcpdump of logging in on port 9090 for cherokee-admin running on the host. http://pastebin.com/m1f637da5 1257431514 M * ryker I filtered out as much fluff as I could. 1257431532 M * ryker this is the command I ran: tcpdump -vvnei eth0 | egrep -v '10022|199\.117\.46\.2|199\.117\.46\.31|199\.117\.46\.144|199\.117\.46\.155|66\.77|snmp|199\.117\.46\.157| arp |ntp|igmp' 1257431555 M * Bertl well, using something like 'tcpdump vvnei eth0 port 9090' might be more effective :) 1257431608 M * ryker ok :) 1257431709 M * Bertl maybe also add 'A' i.e. -vvAnei 1257431975 M * ryker Bertl: http://pastebin.com/m15e9d1de 1257431994 M * ryker that is tcpdump -vvAnei eth0 port 9090 1257432105 M * Bertl okay, now try the same with the guest (make sure to shutdown the server on the host) 1257432133 M * ryker Bertl: ok. 1257432308 M * ryker Bertl: http://pastebin.com/m38fb3d8c Failed one on guest server 1257432469 Q * yarihm Quit: This computer has gone to sleep 1257432470 M * Bertl so the hang would IMHO point to a resolver issue, and the Service Unavailable is definitely something 'userspace' returns (i.e. that's solely configuration dependant not Linux-VServer specific) 1257432489 M * Bertl try if e.g. 'ping www.google.com' works inside the guest 1257432668 M * ryker Bertl: yep, ping www.google.com works fine 1257432688 M * ryker i had lighttpd installed on this guest and it's been working fine for about 6 months 1257432700 M * ryker i just wanted to switch to cherokee 1257432740 M * Bertl well, the 'delay' suggests that it tries to contact something (maybe itself), fails and thus returns with the 503 1257432776 M * Bertl you can try to strace -fF the entire server process, maybe that will shed some light on it, if the software itself doesn't support any reasonable logging 1257432855 M * Bertl google says, CHEROKEE_TRACE should be a good choice (if configured properly) 1257432964 M * ryker thx. I'll give that a try 1257433128 M * ryker Bertl: wow, that dumps out a lot of info 1257433137 M * ryker and i see a connection refused 1257433144 M * ryker let me investigate a bit deeper 1257433831 M * ryker Bertl: I got it to work, but it was a localhost issue 1257433865 M * ryker cherokee-admin uses python and spawns a fcgi server on 127.0.0.1 1257433877 M * ryker I had to change this to my public IP and it worked 1257434402 M * Bertl you could also have simply put ~single_ip in your guest's nflags 1257434448 M * Bertl (see comment about localhost and single_ip an hour ago :) 1257434494 M * ryker sry, i missed that 1257434504 M * ryker i thought that was a compile time option 1257434534 M * ryker sry, i'm fairly new to vserver on centos. i used it on a gentoo for a long time and haven't had this problem for a couple years now. 1257434568 M * Bertl no difference between centos or gentoo (assuming the kernel/tools are the same version) ... anyway, good that it works now :) 1257434672 J * infowolfe ~infowolfe@c-76-105-242-186.hsd1.or.comcast.net 1257434760 M * ryker Bertl: i'm not sure I understand what single_ip does 1257434767 M * ryker does it remove 127.0.0.1? 1257434792 M * ryker or just mirror all network traffic on all interfaces? 1257434986 J * grharry ~root@ppp-94-67-119-158.home.otenet.gr 1257435059 M * grharry Sorry ... I know this is out of topic ... but how can I kick this 2 etc when enter freenode ?? 1257435096 M * Bertl contact freenode services and/or NickServ there, this is OFTC 1257435148 M * Bertl ryker: it basically replaces any binding to 0.0.0.0 with the single IP assigned to the guest 1257435393 M * ryker Bertl: so something that binds to 127.0.0.1 would still be left alone, right? 1257435417 M * Bertl yes, but it wouldn't be allowed to do so, as 127.0.0.1 is not a permitted IP 1257435443 M * ryker ok 1257435540 Q * grharry Ping timeout: 480 seconds 1257435671 Q * SubZero Read error: Connection reset by peer 1257435716 J * SubZero ~SubZero@chello089076140236.chello.pl 1257435751 N * SubZero Guest898 1257435759 J * SubZero ~SubZero@chello089076140236.chello.pl 1257436082 M * ryker Bertl: thx again for all of your help 1257436107 Q * thierryp Ping timeout: 480 seconds 1257436407 M * Bertl ryker: you're welcome! .. see http://linux-vserver.org/Donations if you want to contribute back :) 1257436412 J * dowdle ~dowdle@scott.coe.montana.edu 1257436595 J * yarihm ~yarihm@80-219-11-43.dclient.hispeed.ch 1257437777 P * Alteisen ( $foo ) 1257438671 Q * Guest898 1257438673 Q * SubZero Read error: Connection reset by peer 1257439030 J * imcsk8 ~ichavero@148.229.1.11 1257439115 Q * yarihm Quit: This computer has gone to sleep 1257439320 M * Bertl nap attack ... bbl 1257439324 N * Bertl Bertl_zZ 1257439507 J * dna_ ~dna@p54BC9E60.dip0.t-ipconnect.de 1257439556 N * ensc Guest902 1257439565 J * ensc ~irc-ensc@77.235.160.26 1257439568 Q * Guest902 Read error: Connection reset by peer 1257439901 M * blues Bertl_zZ: anyway - it seems that question 12.7 in FAQ is outdated 1257439906 J * dna__ ~dna@p54BC9E60.dip0.t-ipconnect.de 1257439988 M * blues Bertl_zZ: there is 2 solutions for that: IGNEG_NICE or SYS_NICE. 1257440019 M * blues Bertl_zZ: removing pam_limits.so from files in /etc/pam.d/ solves the problem too 1257440141 A * blues & kid crying :-/ 1257440321 Q * dna_ Ping timeout: 480 seconds 1257440660 Q * derjohn_mob Ping timeout: 480 seconds 1257440777 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1257441201 J * fleischergesell ~fleischer@dslb-088-076-048-058.pools.arcor-ip.net 1257441285 M * fleischergesell hello, im using iptables DNAT portforwarding to make the services running on my vservers accessable to the public - so far so good - but if i try to access my vserver FROM the vservers using the public IP of the server it fails and refuses the connection 1257441301 M * fleischergesell anybody got a clue why and what I can try to fix this? 1257441372 M * fleischergesell to visualize: user@hostsomewhere: nc pub_ip 80 works but user@vserverrunningonhost:nc pub_ip 80 gives conn refused 1257441407 M * fleischergesell it seems as if the packet going from the vserver never reaches the prerouting chain 1257441516 J * thierryp ~thierry@home.parmentelat.net 1257441556 M * fleischergesell is all traffic between host & vserver loopback only? 1257441680 M * ghislainocfs2 fleischergesell: from what i know yes inter vserver is done on loopback 1257441728 M * fleischergesell okay, than it makes perfect sense that the packet never touches the PREROUTING chain 1257443009 M * fleischergesell well, i got no idea on how to tell the packet it should get DNATed to some vserver without using the PREROUTING chain... any idea? 1257443640 M * ghislainocfs2 fleischergesell: i got the issue myself a long time ago but sorry i forgot how i made it work, if bertl goes by he will perhaps have an idea 1257443670 M * ghislainocfs2 then you could test and add your experience to the wiki 1257443688 M * fleischergesell i'd be happy to 1257443698 M * fleischergesell but it seems im figuring it out 1257443760 Q * thierryp Remote host closed the connection 1257444000 M * fleischergesell ghislainocfs2: do you know what the source-ip is when host sends a packet to vserver? 1257444017 M * fleischergesell given there is two interfaces: eth0 with pub_ip & lo on the host 1257444274 Q * davidkarban Quit: Ex-Chat 1257445332 N * Bertl_zZ Bertl 1257445335 M * Bertl back now ... 1257445497 M * Bertl fleischergesell: source IP will be chosen depending on the IPs given to the guests and the target IP, easiest way to figure out _what_ is used in your setup is to tcpdump on the host 1257445677 M * fleischergesell ah, well i just did some tests with netcat and figured out that nc from host to guest uses guests-ip as source 1257445713 M * fleischergesell the thing is I don't know where to put any forwarding if I cant use the prerouting chain 1257445745 M * Bertl there is no 'forwarding' but the packets still traverse input and output chains 1257445788 M * fleischergesell ya, didnt mean the actual chain 1257445791 M * Bertl depending on the kernel, that includes pre and postrouting too, what is your current iptables rile? 1257445794 M * Bertl *rule 1257445826 M * fleischergesell prerouting is not touched as packet never leaves the lo-interface :/ 1257445883 M * Bertl blues: there is a third option, you can simply configure the pam limits correctly :) 1257445940 M * Bertl fleischergesell: don't count on it :) 1257445962 M * fleischergesell right know I do "iptables -t nat -A PREROUTING -d {EXTIP} -m tcp -p tcp --dport 80 -j DNAT --to-destination {WEBSERVER}:80" 1257445979 M * fleischergesell and this rule never gets a count, even if I ommit "-d EXTIP" 1257446002 M * fleischergesell it gets a count obv if i connect from somewhere else, just not if I connect from guest to host_pub_ip 1257446022 M * Bertl your kernel version is? 1257446046 M * fleischergesell 2.6.26-2-vserver-amd64 1257446057 M * Bertl ah, the known-broken debian one :) 1257446061 M * fleischergesell damnit 1257446069 M * fleischergesell :| 1257446114 M * fleischergesell what you mean: it is actually intended that these packets do touch the prerouting chain? 1257446137 M * fleischergesell just I got the damn bugged debian shit? ;-) 1257446158 M * Bertl http://linux-vserver.org/Installation_on_Debian#Issues_with_the_current_2.6.26_Kernel 1257446208 M * Bertl it really varies from kernel to kernel if the packets traverse prerouting/postrouting or just input/output in that case 1257446229 M * Bertl but usually kernels which use input/output allow you to put similar S/DNAT rules there too 1257446266 M * Bertl adding a 'log' type rule entry at the end of each chain usually clarifies the situation 1257446340 M * fleischergesell i did that after noticing that the count for my rule doesn't increase so I do know the prerouting-chain wont get touched whereas the postrouting will 1257446368 M * fleischergesell ah okay, will try to dnat in the input chain 1257446451 M * Bertl in your case, you want the OUTPUT chain :) 1257446517 M * fleischergesell mh, thats weird for me 1257446537 Q * gnuk Quit: NoFeature 1257446565 M * fleischergesell i understand OUTPUT chain for SNAT, but for DNAT? 1257446618 M * fleischergesell but if i think about it.. might actually be sane 1257446622 Q * Piet Quit: Piet 1257447361 M * blues Bertl: very problematic when moving vserver from one machine to another or when you want to fix priority of whole vserver 1257448357 J * Piet ~piet@04ZAAB10Q.tor-irc.dnsbl.oftc.net 1257448515 M * Bertl blues: how so? 1257448969 Q * geb Quit: / 1257449720 J * thierryp ~thierry@home.parmentelat.net 1257449825 Q * fleischergesell 1257450045 Q * thierryp Quit: ciao folks 1257450275 J * unclenick cdlu.netun@cpe-74-66-66-119.nyc.res.rr.com 1257450355 Q * unclenick 1257450374 J * vServer_User ~vServer_U@host90-152-0-26.ipv4.regusnet.com 1257450379 M * vServer_User hi troops 1257450937 J * SubZero ~SubZero@chello089076140236.chello.pl 1257452832 M * yepsen http://linux-vserver.org/util-vserver:InitStyles <- how is a init for gentoo? 1257452904 M * Bertl hmm? 1257452939 M * transacid yepsen: use plain and baselayout-2+openrc 1257453916 J * kbad ~kyle@ip-66-33-206-8.dreamhost.com 1257453959 M * kbad Bertl: I'm seeing this on a few machines that I have and some of them are going unresponsive (the host) any thoughts? 1257453960 M * kbad http://pastebin.com/d505d0730 1257453998 M * Bertl looks like memory pressure ... what kernel/patch? 1257454038 M * kbad 2.6.31.5 from git with vs2.3.0.36.21 & aufs2 1257454053 M * kbad util-vserver is 0.30.216-2 1257454092 M * Bertl that should do the 'right' thing, i.e. kill processes causing the memory pressure (inside the guest over limit) 1257454093 M * daniel_hozac 0.30.216 was released? 1257454150 M * Bertl yeah, and nobody told you :) 1257454166 M * daniel_hozac i guess i've been missing out, huh. 1257454173 M * Bertl nah, that's debian doing an svn release 1257454230 M * kbad I'm not seeing it on all our machines with that kernel though, only a small handful 1257454255 M * kbad have any ideas why it would cause the host to destabilize or is there any other information that I can collect that would help? 1257454291 M * Bertl the question is, what kind of 'destabilization' do you see on the host? 1257454313 M * kbad locks up, unresponsive and has to be remotely powercycled 1257454330 M * Bertl okay, that's bad, let me review that trace again 1257454360 M * kbad I couldn't get sysrq to work either 1257454402 M * kbad if you want the kernel symbols let me know 1257454432 M * Bertl yep, running them through addr2line might sched some light on it 1257454625 M * Bertl you are connected via a serial console? 1257454685 M * kbad yes 1257454707 M * kbad that paste is actually from the buffer 1257454739 M * kbad I'm not familiar with addr2line, lemme look it up 1257454797 M * Bertl add2line -e vmlinux 1257454820 M * Bertl you need the vmlinux not vmlinuz (i.e. not the booting kernel but the built object from the source tree) 1257454832 M * kbad right 1257454917 M * kbad any other options, that seems to be hanging 1257454931 M * Bertl you have to add the addresses now 1257454942 M * Bertl i.e. paste them into it or specify it like 1257454953 M * Bertl add2line -e vmlinux ffffffff810a1ca4 1257454967 M * kbad huh, neat 1257455022 M * kbad http://pastebin.com/d3e495ce7 1257455032 M * kbad bad trace? 1257455102 M * Bertl nah, looks more like kernel missing DEBUG_INFO 1257455119 M * kbad ahh 1257455181 M * kbad I guess I could build one and throw it on one of the problem hosts 1257455187 M * kbad does that sound like the next step? 1257455202 M * Bertl yes, definitely, also enable some other debug and trace features 1257455265 M * Bertl daniel_hozac: what do you think, should we force DEBUG_INFO on in the future? 1257455288 M * daniel_hozac no 1257455309 M * daniel_hozac without proper stripping and such, it increases the module sizes by orders of magnitude. 1257455320 M * Bertl ah, okay ... 1257455367 M * kbad daniel_hozac: does adding frame pointers bloat it much? 1257455375 M * daniel_hozac and i don't think most people do strip their kernel modules. 1257455387 M * Bertl kbad: nope, and it's a good idea for debugging 1257455435 M * Bertl daniel_hozac: would be a good idea to do the stripping on install_modules 1257455465 Q * bonbons Quit: Leaving 1257455478 M * daniel_hozac yeah 1257455519 M * kbad where is debug_info I can't seem to find it in my .config or menuconfig 1257455547 M * Bertl daniel_hozac: btw, I decided to put the checks in the various dev_* helpers which fetch the device by index/name/etc 1257455586 M * Bertl tests so far seem to give the desired results ... let me upload a patch for you to see what I mean 1257455647 Q * manana Ping timeout: 480 seconds 1257455852 M * Bertl daniel_hozac: http://vserver.13thfloor.at/ExperimentalT/delta-devhide-fix01.diff 1257456181 M * kbad Bertl: any specific tracers? 1257456382 M * Bertl http://paste.linux-vserver.org/13599 1257456409 M * Bertl also the soft watchdog might be a good choice, so that we can get a trace when it 'hangs' 1257456656 Q * vServer_User 1257457070 M * kbad ok kernel built 1257457072 M * kbad lemme get it rockin 1257457249 M * daniel_hozac Bertl: okay. looks fine to me. 1257457271 J * manana mayday_780@84.17.25.144 1257457394 M * Bertl Guy-: could you give http://vserver.13thfloor.at/ExperimentalT/patch-2.6.31.5-vs2.3.0.36.22.diff a spin and let me know if that fixes the lmhostid issues for you? 1257457411 M * Bertl daniel_hozac: thanks for checking! 1257457459 Q * larsivi Remote host closed the connection 1257457525 M * Guy- Bertl: in principle yes, but not today :) thanks for looking into it though 1257457558 M * Bertl no problem, thanks for spotting and reporting it! 1257457589 M * Guy- Bertl: in fact, if the ioctl now returns ENOENT or whatever it does if no such interface exists, the flexlm license manager may be almost completely broken in vservers :) 1257457620 M * Guy- (but I agree it's the correct behaviour) 1257457631 M * Bertl it'll still be fine when you have that device assigned to the guest and/or use the preload library 1257457651 M * Guy- it looks up an interface called 'xp0' before any others 1257457663 M * Guy- so it'll always be a valid workaround to give it an interface called xp0 1257457688 M * Guy- (even if I don't want to give it the host's eth0 and don't want a network namespace) 1257457715 M * Bertl yeah, generating uique IDs based on MAC addresses is one of the weirdest ideas I ever heard of, but it's quite common for license managers :) 1257457775 M * Guy- it's another good example of copy restrictions that only cause pain for legitimate users - if I wanted to use an illegal copy, I could use a cracked version and be done with it 1257457807 M * Bertl on linux, not even a need to do so, as you can set the MAC to anything you like :) 1257457835 M * Guy- fwiw, you can even do that in Windows, at least with many NIC drivers 1257457860 M * Guy- so yes, it's pointless as a copy protection measure but inconveniences the customer who paid for it 1257457886 M * Guy- free software ftw 1257457899 M * Bertl as almost all of those solutions ... okay, let me know how it goes when you get to test it 1257457927 M * Guy- I shall 1257458306 J * uva_ bno@118-160-161-100.dynamic.hinet.net 1257458743 Q * uva Ping timeout: 480 seconds 1257459158 Q * Guest205 Ping timeout: 480 seconds 1257459202 Q * transacid Ping timeout: 480 seconds 1257459226 J * Genghis ~Genghis@ph34r.my.d-n-s.org.uk 1257459258 N * Genghis Guest934 1257459671 Q * kbad Quit: Leaving. 1257460294 Q * ghislainocfs2 Quit: Leaving. 1257460407 J * derjohn_mob ~aj@d004209.adsl.hansenet.de 1257460418 Q * SubZero 1257460431 J * transacid ~transacid@transacid.de 1257461145 J * jla ~jacob@67.88.119.7 1257462026 M * jla What does --pkgbase/PKGBASE do, what is it for? 1257462105 M * daniel_hozac external package management. 1257462122 M * Bertl i.e. where to place the necessary info/packages/etc 1257462146 M * jla What would I use it with? I've searched all the archives and web site trying to get a vision on it 1257462177 M * jla my current guess is for rpm's --relocate based on some other "internalize" concept I also dont get yet :) 1257462203 M * jla at first I was guessing it was where package files were cached 1257462204 M * Bertl you know what external vs internal package management means, yes? 1257462262 M * jla I've kinda got the concept from one page giving an example with that "internalize" command. One way you run the command outside the guest and stuff the output into the guest's filesystem, the other way is's like a "normal/real" machine 1257462275 M * jla thus requiring a relocatable package system 1257462279 M * jla like rpm --relocate 1257462288 M * daniel_hozac no 1257462290 M * jla or maybe extract to temp, stuff files 1257462295 M * daniel_hozac relocate is something entirely different. 1257462296 M * jla glad I'm way off :) 1257462309 M * daniel_hozac --root is more like it. 1257462337 M * Bertl the point is, to keep the package management on the host (externalized) or within the guest (internalized) 1257462372 M * Bertl i.e. all the information about installed packages as well as the tools required to manage them can be _outside_ the guest 1257462435 M * jla like the rpm or apt database? 1257462440 M * jla of what's installed 1257462446 M * Bertl exactly 1257462485 M * jla would I then need a pkgbase per guest if they had different package needs? 1257462492 M * daniel_hozac no. 1257462507 M * daniel_hozac it's a base directory, i.e. each guest gets a different subdirectory below that. 1257462606 M * jla so package database goes to .pkgbase/guest instead of /vserversroot/guest/var/lib/pkgbase 1257462629 M * jla or whatever they call it on the end like apt or rpm 1257462638 M * Bertl (or wherever it would go with whatever package management is used :) 1257462753 M * jla it wouldn't stop a person with root in the guest from installing apt or rpm would it? and if they did and then from the host we internalized that could be a mess? 1257462796 M * Bertl yes, the externalized management is more suited for setups where guest and host are administrated by the same person/group 1257462843 Q * dna__ Quit: Verlassend 1257462869 M * jla thank you for clearing that up :) 1257462900 M * jla you guys do some custom trickery to wrap a bunch of packagemanagers for that, or does it only work with a few like rpm/yum? 1257462921 M * daniel_hozac currently only rpm is supported. 1257462934 M * jla maybe the --relocate flag helps? :D 1257462938 M * daniel_hozac (and, by extension, yum) 1257462997 M * jla good to know. I've been using vserver for over a year. pretty awesome stuff, but still feel so much like a consumer 1257463067 M * Bertl well, that's what the channel is for (information), feel free to ask whatever you want to know about it :) 1257463112 M * Bertl sidenote: with over a year you are pretty 'new' to it compared to the history of Linux-VServer :) 1257463128 M * jla true there 1257463313 Q * dsoul Ping timeout: 480 seconds 1257463375 J * dsoul darksoul@insomniac.pl 1257463750 Q * imcsk8 Quit: Leaving