Server :: General Authentication - Significant Delay After Pass Entered / PAM_unix
Feb 23, 2010
My first post here. I've been using Ubuntu Server edition (Hardy) happily for some time now. I use sudo regularly during configuration of new services. It always works/authorises within seconds, however, it recently became very slow, to the point of being nearly unusable. It turns out that ANY authentication request stalls in this way - in in a plain shell, X or Gnome, via VNC (and even FTP seems affected). In /var/log/auth.log I noticed a regular working pattern like this:
[Code].....
Unfortunately now, these records look the same, but the timestamps are much wider apart (minutes). The problem occured after running synaptic and dpkg-reconfigure obm-conf. I was trying to remove 'OBM' [URL] completely as I'd messed up (just OBM) while trying to fix a minor-looking misconfiguration.
I've carefully analysed which files were altered (modified) during my changes (as the problem survives a reboot), but there are no likely-looking candidates in the list (I deleted some blank lock files). I also did a wireshark capture during an authentication session and noticed that it doesn't appear to be looking off the server for information (I had likewise-open installed, but the problem exists whether that is running or not). Research in the forums hinted that this might just be as simple as a bad /etc/hosts file - but /etc/hosts has not been modified. I may have a flawed DNS, but I get fast enough responses to nslookup and can access Internet etc.
I'm running PCLinuxOS with no other OS installed. Thought I'd have a go at chainloading. just for interests sake.So, I opened Konsole, entered su and entered my root password. The command line turned red so, I imagined all was well. But when I then keyed in '/boot/grub/.menu.lst' I was refused access.
how to do a small script that checks if what the users entered is an ip address or not? Also possible to do the same but instead of an ip address is a netmask. For example the user entered: 192.168.0.1 so its true this is an ip address, if it doesn't look like an ip address then is false And second example, the user entered: 255.255.255.0 so the script returns true, and if its not a netmask the script returns false.
I already have two other threads open for a dysfunctional system. Thread 1: Boot hangs at logon sound - no video signal - appears drive is still working [URL]..Thread 2: Where are boot log files and how I can see those? [URL]...Basically both of those are for not able to complete the boot (hanging up in the middle of booting) in Karmic. I had Karmic since it came out but suddenly about two weeks ago just stopped booting. Except for the battery and resetting BIOS, I had checked all the hardware - and all appear to be okay. Following a tip from x1a4, I started looking into logs (admittedly not knowing exactly what I am looking for) and I ran into auth.log. Following is the code. The highlighted part is confusing (to me) and I am wondering if this may cause to stop booting (or completing the boot process) since I was using auto logon. I do not know if there is anything else (logs) you may need, please advise and I will provide those.
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.
During a recent request to install fedora 11, I ran into an interesting problem. It seems that between fedora10 and fedora11, the developers switched from fdisk, to parted for creating the initial pre-mke2fs partition table creation.
It looks like the implementation of this is broken, as it's writing the partition table with overlapping cylinder boundaries. While this can sometimes be ignored, it can in certain cases cause significant data corruption.
On an installation I took it through, using the latest installation media, in both manual & automatic partition creation, the layout looked like this:
The last two partitions turn out fine, for some reason. However, those two partitions should not have overlapping cylinders. After my very first installation, the system was completely unbootable, and not even fsck wouldn't rescue it. If this is possible, then that means that a system that's been online for months or even years could simply drop out of functionality simply due to a byte or two of system-critical data falling on that last cylinder. Considering that a lot of the time kernel data ends up on /dev/sda1 (commonly the /boot partition), this is something that should not be ignored.
When I am trying to run the Xserver using the command startx I am getting the below mentioned error
xauth: creating new authority file /oracle/oracle10g/.serverauth.22555 Fatal server error:PAM authentication failed, cannot start X server. Perhaps you do not have console ownership?
After GRUB 2 comes up (I'm running Ubuntu 10.10) and I choose the OS to boot, there is about a 5 second delay where nothing appears to happen after I make the selection -- no disk activity. It happens consistently every time I boot. Again, this is after I choose the OS to boot, so it shouldn't have anything to do with the standard delay to allow me to choose the appropriate OS.Is there a good way to troubleshoot this and determine what is causing the delay?
I'm in normal mode then I press "ayiw to yank a word into register a, gvim shows the "ayi as I type it near the bottom right corner. Is it possible to get this real-time display in console vim, if so how?
I just got a new laptop for work that dual boots with Windows 7 and Ubuntu 10.10.Normally, I can use an sslvpn connection to log onto my office servers to transfer data back and forth through Windows. However, I do my actual work on Linux, which the computing services department at the university where I work has not set up a sslvpn connection method for.I got the JRE set up on Linux and loaded up the sslvpn connection when I was asked for my sudo password. After following some other instructions, I created a different password to act as root.I entered that password over and over again and now I'm not able to attempt to connect again by trying another password.
I was tired and went to delete a directory and its content and forgot to specify the directory name, i entered: rm * -R Now lost all what was in my home directory (((
is it possible to create a dhcp server which will ask for authentication from client before providing ip to that client ? i am using rhel5 os. i have configured dhcp server and it is working fine.
I'm fairly new to Linux and very new to Squid and am having authentication issues! I am using Oracle Enterprise Linux (which is basically Red Hat without the branding) and wanting to use Squid Proxy Server for web access with authentication to Active Directory. I've found a number of articles about this online and all of them say to use auth program squid_ldap_conf which should be in /usr/lib/squid/. I don't have a squid directory in /usr/lib for starters and my squid binaries are in /etc/squid but there is no squid_ldap_conf in there either. I have installed the latest version of Squid (3.0) to see if that helped but I still cannot find the authorisation program.
I need to set a delay between sending two mails (say 1 second). I have a shared server so delay for specific domain will be more usefull(aaa.com - 2 sec, bb.com 1 sec etc).
I have debian lenny, when I run an application is slow to load (example: iceweacel open and it takes, I can hold him iceweacel work normally), I tried to change from gnome to another and the same thing, went down some services (samba, squid) and nothing. I open a terminal and it takes, I want to duplicate it and do it fast. From a terminal without X (tty1) with root run mc and moves quickly, also run as root "sudo mc" and takes to boot.Any action done with sudo it takes to run.Can not be what it takes to make starting the applications, not the PC because it is new, and from one moment to another I began to pass this
I am using centos 5 for squid. I have enabled delay pools in it. The scenario is I have two clients machines have IP address 192.168.1.200 (The Manager) and 192.168.1.201 (Manager Assistant). The issue is I assign 512kbps to the Manager IP address and 128 to his assistant. When either of them download it start from 60 KB/Sec they finally get 11 KB/Sec which is wrong. If I disable the assistant delay pools settings then the Manager get 60KB/Sec continuously. By enabling both squid does not show any error.
Following is my line of code taken from squid.conf Code: acl our_512kbps src "/usr/local/squid/etc/our_512kbps" acl our_128kbps src "/usr/local/squid/etc/our_128kbps" http_access allow our_512kbps http_access allow our_128kbps http_access deny all
I would like to ask whether there is a way to postpone or delay message processing on postfix mail server. E.g. to postpone message transfer/relay for 5 minutes to be able to dequeue mis-targeted emails.I couldn't find anything because these keywords mislead to performance problems.
How do you configure proftpd so that once a user has failed to login and reached the MaxLoginAttempts. That they can not retry logging in for another 4 hours ?
I'm trying to lower consumption of my server/HTPC. After wakeup from pm-suspend server/HTPC is ready to use in 1 or 2 seconds .For example if I suspend it while watching movie in KODI, after resume movie starts playing instantly. But some services (SSH and SAMBA) are not running. I thought it was network problem so I change configuration to static (not DHCP). SAMBA and SSH starts like 15 seconds after wakeup.
Sendmail on the RHEL server is set up to forward email to company internal mail server (via SMART_HOST). This has worked fine for years.
From the server, doing a sendmail dave@company.com, the message is sent immediately. No problems - and no change here.
From the server, doing a sendmail roberto@compania.com (compania is the part of the company in Mexico) causes a 2m40s delay before sendmail does anything.
No packets are sent until after 2m40s wait. Message is always sent after the delay. It is not clear what the server is waiting for.
For *each* user in the @compania domain specified in sendmail command, an additional 2:40 is imposed before email is sent. Email is always sent after the n*2:40 sec delay.
Sending email to some_name@AnUnusedDomainNameThatDoesntExist.com is processed immediately - with a bounce email received in a couple of seconds.
Q: How do I determine where this 2m 40s delay is introduced?
We have a squid-proxy configured with 1 delay pool to limit the bandwith to 6M. I have to create a kind of exception for a specific remote host for which we want to reserve 1M which is not included within the 6M. How should I do that? Here is our actual configuration of the delay pools
I am not parsing on a webserver so is it possible to have both
#! /usr/bin/php & #!/bin/bash
in the same script? Alternatively, I have a current bash script that I need to get some variables from mysql and not sure how to get mysql results in bash:
Quote:
mysql -h server.net -u username1 -paaa -e "USE squid; SELECT email, usern FROM TABLE WHERE blah blah;" emailadd="resultfrom above" usern="resultfromabove"
