General :: Pam_mysql_check_passwd() Returning 6 - Pam_mysql Authentication?
Aug 2, 2011
I'm having trouble with my pam_mysql authentication.'ve double checked my md5 password but its correct.What should I be looking at? My /etc/pam.d/system-auth is completely identical to my other server (that works), but this one is somehow reporting that my password is wrong.I've bolded out the questionable output.Here's my error log while trying to connect:
Code:
Aug 2 16:16:21 la2 sshd[8006]: pam_mysql - option verbose is set to "1"
Aug 2 16:16:21 la2 sshd[8006]: pam_mysql - option db is set to "hidden"
If I am running a script, let's say a install script. Is there a way to make Su repeat authentication rather then just returning "Authentication failed" and continuing the script?
I'm interested in using pam_mysql or pam_ldap for authentication with my vsftpd server on CentOS. Does anyone know if, when pam_mysql connects to a remote database, whether or not the connection is, or can be, encrypted with SSL or some other means? While I know that passwords are hashed, I'd rather not have my usernames or my pam-related SQL queries travel in the clear.
If pam_mysql DOES communicate only in the clear, what about pam_ldap? I'm more familiar with LDAP than pam_mysql, and have always secured my conversations with LDAP servers when using an LDAP client or my own LDAP connection libraries.
On Ubuntu server 10.10, with a relay smtp server with authentication via postfix; I keep getting 535: Incorrect authentication data. I'm sure my username and password is correct. Heres how I set up postfix: I created a file called smarthosts.conf in my /etc/postfix/ directory that contains the following:
[Code].....
my server uses plain text authentication on port 25. I would like to use security like SSL, but this particular server is unsecured.
I need to make a choice on what authentication protocol I want to use for Authentication and Authorization. I was looking at Radius and then literature suggested that Diameter was a better protocol. Keep in mind I need this on a hetrogeneous setup ( linux & windows together). Diameter seemed like a good fit until I discovered that the open source code no longer seems to be maintained ( C/C++).
I was also looking at Kerberos as an option though there is alot overhead with the server. SSL/TLS or EAP? I am looking for simple but secure and am new at the security protocols.
I have a problem with ldap client authentication in ubuntu. I am using rhel5 as openldap server and I configured ubuntu as client, when I am trying to login the following message is coming."su: Authentication service cannot retrieve authentication info. Sorry" But when I do search through "ldapsearch" command output is coming without any errors, Can anybody explain what would be problem.
i'm trying to make a backtrace of nautilus (as described here: URL...to fill a bug report but it doesn't work.after i "attach PID" to the nautilus-process and do "continue" it's not possible with ctrl-c to return to the (gdb)-prompt. the wiki-page above says:"If the program hangs but doesn't crash you can press ctrl+c in gdb while the program is frozen and then continue with the next step".
I configured Dell Latitude D620 to go to sleep mode (suspend to ram) when i close the lid. It suspends ok and then I open the lid. It wakes up ok, but few seconds after goes to sleep again! So I need to press the power button to force it to wake up.
Sleep button works just fine. /etc/acpi/sleep.sh from console too.
I noticed that it always goes to sleep again when I wake up laptop by opening the lid. For example if I run /etc/acpi/sleep.sh, pm-suspend or press the sleep button laptop goes to sleep, then i close the lid and when i open it back - the problem occurs.
I am facing the problem of calling cfitsio library from C program in linux. when i am givingthe option "gcc -o CreateLevel1DataFitsFile CreateLevel1DataFitsFile.c -lm -lcfitsio "i am getting following error
/usr/lib/../lib/gcc/i386-redhat-linux/3.4.4/../../../libcfitsio.a(getcolj.o)(.text+0x7065):/backup/Astrosat/learning/cfits/cfitsio/getcolj.c:3098: more undefined references to `__xtoll' follow
I am developing a script to automate some database setup, but I have an issue I cannot figure out. I prompt for some input with whiptail and select a default if nothing is entered. However, if you do not enter anything, it normally returns 0, but this time it is returning an empty value.
Bit of a strange one - locate is not finding any files ?? I'm running xubuntu karmic and 'locate' is actually 'mlocate' so doesn't allow the -u switch.
I don't know what I'm missing or what I did to break it, but most screensavers that I have end up returning this error message: "X Error of failed request: BadRequest (invalid request code or no such operation) Major opcode of failed request: 135 (GLX) Minor opcode of failed request: 19 (X_GLXQueryServerString) Serial number of failed request: 16 Current serial number in output stream: 16" This is also the output of glxinfo. Did I somehow break something in my graphics driver?
Im constructing a menu for a program using case. I have all my normal input options mapped out but i want to have a * ) option so that if something else is inputed, it displays "Incorrect input" and then resumes the normal menu function. How do I do this? Ideally, I'd like it to display "Incorrect Input" and then accept more input for the menu.
I have a linux box, which was given to me by a company with 3 dedicated IP addresses. My goal was to setup ssh tunnels on all three dedicated IP Addresses.
Via Putty, i am successfully able to create a tunnel to the box's on IP1 IP2 and IP3 - that is i can make incoming connections on all three IP's.
However, my problem lies when i attempt to USE the tunnel. AIn all instances, regardless of the IP i actually tunneled into the box with, it returns IP1 as an outgoing connection, i.e if i setup IE to use the tunnel, it will return my IP as IP1, even though i have created the tunenl via IP2/IP3.
What id liek to accomplish is that when i create a tunnel via IP1, then outgoing connections go via IP1, and when i create a tunnel via IP2, then outgoping connectiosn go via Ip2 etc.
After discussing this with a colleague of mine, he infomed me he had the same problem at one point, and that some configuration on the box was required - however he was unable to recall.
I'm building a very small netbootable client, using a stripped down custom kernel and busybox v.1.18.4 (stable) with the embedded ifconfig and udhcpc applets
after startup, the initrd is loaded, and the proc, sys, dev and pts folders are properly created and mounted.
/proc/net/dev properly contain eth0 as an ethernet device.
in the init script I execute a background call to udhcpc passing the dhcp.script properly to udhcpc, and dhcp.script is then executed when the bound event occurs.
what I cant get past is udhcpc is passing back the interface evironment variable as etho0 (thats, 'eth' lowercase 'o' and a zero '0', when it should be passing 'eth0', thats 'eth' and zero '0')
I know I could hardcode the eth0, but i was hoping for a somewhat smarter system.
I have a situation where I have two DNS servers - a BIND one for our Unix hosts and a Windows Server DNS one for the Exchange and other Windows hosts. I set up the BIND server to slave from the Windows one, which it seems to do just fine.
Windows DNS has entries like "_gc._tcp.domain.local" in it, which are needed for domain logins, etc.
However, when I try to use nslookup on those entries, it fails when querying the BIND server. When I do a domain transfer with dig, I get all the answers I expect.
This failure means that the Windows boxes can't query the BIND server for the answers they need. By default, from DHCP, it is the BIND server they get, so that's a big problem. If I manually switch their DNS to the Windows server, it works perfectly.
I have a secondary slave dns server setup, it has transferred all the records from the primary master.But when I run a few checks from outside using tools on the net it says:Tried to fetch SOA record for domain, but DNS server ns2.domain.com returned error code Refused.I disabled iptables, and tried again, same result. So I think it is something in the named.conf file? I appreciate any extra set of eyes on it. Just feel like I'm so close!
I installed the stable version of Debian last night (Debian GNU/Linux 6.0 l) and whenever I come out of suspend, the video card's fan stays at full speed. When I do a regular reboot, this doesn't happen.
Normal behavior is for the fans (cpu and gpu) to go to full speed for a few seconds and then settle on a slower (quieter) speed.
The hardware involved is: Dell Vostro 400 01:00.0 VGA compatible controller: nVidia Corporation G92 [GeForce 8800 GT] (rev a2)
The install is vanilla (as in "I haven't messed with drivers or configuration files - yet"). Also, keep in mind that I'm a newby - I've just switched to Debian after about two months of Ubuntu and am not too familiar with where everything is or what commands I need to do what, so instead of telling me to change the sync parameters in the nvidia configuration, I need to know exactly what parameter in exactly what file.
Every time I leave runlevel 5 (shutdown, reboot, init 3 or whatever) and then return to runlevel 5, I lose sound and have to delete and then re-edit sound card config in Yast Sound module to restore it.When returning to GUI environment, the sound card appears OK in Yast but I get no sound from any of my multimedia apps - Amarok, VLC, Firefox all go quiet. Strangely, it still plays the KDE intro jingle when KDE starts up.To restore sound I click delete in Yast and I'm prompted with the following KDE dialogue:
Removed Sound Devices KDE detected one or more internal sound devices were removed. Do you want KDE to permanently forget about these devices?
I do like a lot of the changes they've made in Lucid (especially the aesthetic ones), but there is one change they made that I don't like, and I can't figure out how to get around it. I have Ubuntu set so that, after 10 minutes or so of inactivity, my screen is blanked, in order to save power. In Lucid, though, whenever I come back to my computer after the screen has been blanked, it requires me to entire my password in order to resume use of my computer. Is there any way I can turn this off?
This next question is unrelated, but I do still feel the need to ask. Why, exactly, did they move the minimize/maximize/close buttons from the top-right corner of the window to the top-left? I know that it probably feels natural to Mac users, but every previous version of Ubuntu has had these buttons in the top-right corner.....why on earth did they feel the need to change it? I've been using Ubuntu for a while now, and it's kind of annoying. Surely most people who've been using Ubuntu for a while were thrown off by this, as well? I doubt that there is any way to change it, but does anyone know if there is?
After returning from suspend to disk, RF devices (wlan and bluetooth) are disabled. Running lspci indicates that no wireless network device is available, and no drivers are loaded. Furthermore, the LED on the front of the laptop remains off - this is always on when RF devices are available. There is an Fn-F1 function key to toggle wireless, but this does not function to either enable or disable.
Rebooting the system does not re-enable the RF devices, and there is no option in the BIOS. The only workaround I have at present is to boot into windows vista and use the Fn-F1 function key. This pops up a small hotkey utility with the option to disable/enable the wi-fi and bluetooth (independently). Upon confirming the dialog, the RF device light immediately illuminates. Rebooting into fedora and wireless devices are working once again.
I am running kernel 2.6.31.12-174.2.3.fc12.x86_64, with the latest compat-wireless drivers. I can provide much more debug information (dmidecode, acpidump, boot logs, lspci etc...), I am just unable to determine what is relevant
# rfkill list 0: hci0: Bluetooth Soft blocked: no Hard blocked: no
