Bug 18221
| Summary: | linuxconf-web does not work | ||
|---|---|---|---|
| Product: | [Retired] Red Hat Linux | Reporter: | Tim <bodysurf> |
| Component: | linuxconf | Assignee: | Nalin Dahyabhai <nalin> |
| Status: | CLOSED DUPLICATE | QA Contact: | |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.0 | CC: | darrel, karlheinz, marwan, matzilla, milan.kerslager, redhat, sjledet, sysadmin |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | i386 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2001-04-09 10:04:07 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
try in /etc/xinet.d/linuxconf-web to change the line "disable = yes" to "disable = no" Changing the "/etc/xidentd.d/linuxweb-conf" doesn't do a thing. I'm also having problems with linuxconf --http. I cannot get it to run on the command line or in a browser. When I do a linuxconf --http --debug, I get a segment fault(core dumped). hello, I also have the problem I set disabled to no in /etc/xinetd.d/linuxconf-web restarted xinetd I use telnet localhost 98 to test each time i type a line I'm disconnected unless I have another telnet open If I use linuxconf --http from a command line linuxconf automagically runs in debug mode and open a socket on port 8000, where I can successfully connect I've traced the problem to dialog/html.cc in html_get the command ok=cmd->listen(timeout*60) never returns (this function is defined in dialog/cmdsock.cc) by the way, the correct syntax is linuxconf --http --debug 0|1 it is also possible to use --name or --port I've send a little patch to the maintainer, Jack Gelinas, so that it no longer coredumps for this trying to access linuxconf-web trough port 8000 doesn't work for me... is suppossed to work in this broken release ? amazing that nalin hasn't tryied linuxconf-web before the release... to connect to port 8000 in debug mode, you have to start linuxconf --http in a shell before (so you completely bypass xinetd) you can do a netstat -a to show the open port for the linuxconf-web problem, it seems that indeed xinetd is bugged it should handle, as inetd does, the control socket to linuxconf but instead it handles only the data part ! Oct 13 12:49:32 orpheus xinetd[407]: START: linuxconf pid=7222
from=203.12.234.232
Oct 13 12:49:32 orpheus xinetd[407]: EXIT: linuxconf signal=11 pid=7222
Unless I'm mistaken it appears linuxconf is segfaulting?
I added the following to the linuxconf-web config to generate that:
log_type = SYSLOG daemon
log_on_success += EXIT
We've fixed this in xinetd, coming as an errata soonish. The updated xinetd does not fix the problem. I still can't use linuxconf-web - when starting it from the command line /sbin/linuxconf --http I only receive an "ending html". the updated xinetd works for me :-) I can connect to http://localhost:98 I've set disabled= no in /etc/xinetd.d/linuxconf-web I'm using linuxconf 1.21r6 (compiled from source) It ididn't work with the orginal RPM. I updated to 1.21r1 and it works now The updated xinetd package did not fix the problem. Now I get: Oct 23 11:38:53 orpheus xinetd[3449]: warning: can't get client address: Invalid argument Oct 23 11:38:53 orpheus xinetd[3449]: refused connect from 0.0.0.0 Oct 23 11:38:53 orpheus xinetd[3449]: FAIL: linuxconf libwrap from=0.0.0.0 Oct 23 11:38:53 orpheus xinetd[3449]: file descriptor of service linuxconf has been closed Oct 23 11:38:53 orpheus xinetd[3449]: select reported EBADF but no bad file descriptors were found After that it just refuses connections from anywhere. When I install linuxconf 1.21r8 (and devel) and Xinetd update, Got a error message in /var/log/messages: Oct 30 10:24:16 mail xinetd[555]: file descriptor of service linuxconf has been closed Oct 30 10:24:16 mail xinetd[555]: select reported EBADF but no bad file descript ors were found After install linuxconf-lib-1.21r8-1.i386.rpm and linuxconf-util-1.21r8- 1.i386.rpm, it seems like working fine so far. Linuxconf-web was down every 15 to 20 minute, now it has been running for an hour, and no error at all. Unfortunately when linuxconf dies (time-out or kill), xinetd occur EBADF and after that it rejects any access to port 98. I also tested this with linuxconf 1.22-1 and xinetd 2.1.8.9pre12-3. Actually I use inetd from RH6.2 for linuxconf-web and xinetd for other services. How do we reopen this? It is obviously not resolved?! Nov 30 17:18:38 smail xinetd[31666]: warning: can't get client address: Invalid argument Nov 30 17:18:38 smail xinetd[31666]: START: linuxconf pid=31684 from=0.0.0.0 Nov 30 17:18:39 smail xinetd[31666]: EXIT: linuxconf status=255 pid=31684 Nov 30 17:18:39 smail xinetd[31666]: file descriptor of service linuxconf has been closed Nov 30 17:18:39 smail xinetd[31666]: select reported EBADF but no bad file descriptors were found NOTE: I'm not calling it from 0.0.0.0?! there's a fix for xinetd which solves the problem. see http://www.solucorp.qc.ca/howto.hc?projet=linuxconf&id=42 Still stuffed for me. None of the above fixes solve any of the problems. Still broken. Posts to Jacques Gelinas also give no Joy. Linuxconf will work from the local machine but never as web interface. Linuxconf looks for a linuxconf-web script when initializing the web interface from the command line linuxconf. This script does not exist. When booted, RH7 initiales a listening and waiting port 98. This port dies unceremoniously as soon as I attempt to access it. I am running DHCP, Samba (SWAT does work fine), print server, ipchains. Everything works fine except linuxconf-web. The error message when I attempt to rouse linuxconf over the local net is: Jan 4 10:05:25 grippo xinetd[537]: warning: can't get client address: Invalid argument Jan 4 10:05:25 grippo xinetd[537]: refused connect from <no address> Jan 4 10:05:25 grippo xinetd[537]: linuxconf service was deactivated because of looping Jan 4 10:10:00 grippo CROND[3747]: (root) CMD ( /sbin/rmmod -as) The 'invalid address' repeats 10 times before it terminates the service. my netstat -nr is: Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 255.255.255.255 0.0.0.0 255.255.255.255 UH 0 0 0 eth0 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 65.24.84.0 0.0.0.0 255.255.252.0 U 0 0 0 eth1 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo 0.0.0.0 65.24.84.1 0.0.0.0 UG 0 0 0 eth1 I have hosts.deny set to all:all and hosts allow to ALL: 192.168.1.0/255.255.255.0. I have not set any explicit routes and have installed RH7 clean with only using the firewall howto to get things running. works perfectly for me : RH7 + all errata xinetd rpm with fix from linuxconf web site (see url above) because errata from RH contains only a partial fix in /etc/xinetd.d/linuxconf-web you have to set disable= no check that the server path is good (else you have a warning in /var/log/messages) restart xinetd : /etc/init.d/xinetd restart check logs tries http://localhost:98 I use linuxconf 1.23 This is still not resolved. It works fine as long as you don't have an active /etc/hosts.deny file, but when hosts.deny is active, the following appears in /var/log/messages: Jan 7 19:57:44 linux xinetd[494]: warning: can't get client address: Invalid ar gument Jan 7 19:57:44 linux xinetd[494]: refused connect from <no address> To fix the above error if you are using a hosts.deny/hosts.allow file, you can add linuxconf: ALL to the hosts.allow file. This makes it work. Thanks for the hosts.allow comment. I knew I wasn't a total idiot. I feel this should be fixed a little different as this may pose a security risk. It sounds like it is an xinetd problem... The update xnetd(xinetd-2.1.8.9pre11-1.i386) stil not work When i start or reload xinetd "linuxconf-web" the service stay up for 1 thought 3 minutes...and down after... i have upgraded packages linuxconf and xinetd.. and trouble continues... any solution...???? Go to http://www.solucorp.qc.ca/linuxconf and install the xinetd which has the "wait" patch xinetd-2.1.8.9pre11-1wait.i386.rpm cause the xinetd from RedHat doesn't fix it. ftp://ftp.solucorp.qc.ca/pub/misc/xinetd-2.1.8.9pre11-1wait.i386.rpm What is the secret combination of versions to get this to work. Im running Linuxconf 1.25 - xinetd 2.1.8.9pre11-1wait and I still get the error Executing: /etc/rc.d/init.d/linuxconf-web: start * sh: /etc/rc.d/init.d/linuxconf-web:: No such file or directory * return 127 Does anyone have this working??? Thanks The bug in linuxconf-web was fixed only when i upgraded xinetd to xinetd-2.1.8.9pre11-1.i386 and remove line disable = yes from linuxconf-web and configure the linuxconf network access. After this restart xinetd.conf. And it should be worked. *** Bug 14986 has been marked as a duplicate of this bug. *** *** Bug 33608 has been marked as a duplicate of this bug. *** I upgraded to xinetd-2.1.8.9pre14, the linuxconf now start (via xinetd, the old xinetd from RH 7.0 simply bomb out) but nothing more happened. linuxconf --http --debug still gives Segmentation fault. Manual connectin to port 98 by telnet I can see that linuxconf write back nothing after a passed to it GET / HTTP/1.0 (connection was closed after submiting query). I tryed to upgrade to linuxconf-1.24r2-10. I had to update more packages (mount, util-linux, pam, etc) and now linuxconf do not close the open connection but gives nothing to browser. When trying WWW acces through WWW client I can see that xinetd kill linuxconf after 48 sec but there is still nothing usable on the screen. My DNS and other services are Ok, I have pernament acces to the Internet and I have no other troubles with anything. *** This bug has been marked as a duplicate of 23560 *** |
Just installed RH 7.0 and can't get linuxconf access via HTTP. Works fine in text and in GUI mode. I try to activate it via linuxconf text mode -> Control -> Control Service Activity -> linuxconf linuxconf is set to "manual". I set it to "automatic" and that registers. I try to "start" it next at the same spot and it returns a: Executing: /etc/rc.d/init.d/linuxconf start > Running Linuxconf hooks: [ OK ] but when I hit the "accept" button and go back a level, linuxconf is not running. I tried getting linuxconf-web to run also, which was set at "automatic". I tried changing it to "automatic" and I get a: Executing: /sbin/chkconfig linuxconf-web: reset * error reading information on service linuxconf-web:: No such file or directory * return 1 ! Switch service linuxconf-web: to Automatic I tried to "start" it next and I get a: Executing: /etc/rc.d/init.d/linuxconf-web: start * sh: /etc/rc.d/init.d/linuxconf-web:: No such file or directory * return 127 I don't even see a '/etc/rc.d/init.d/linuxconf-web" file in the RH 7.0 RPM.