GSP will not let me login to root

[Timothy P. Jackson]

Hello Everyone,

I have an issue when trying to access the console login from the GSP.

I telnet into the GSP which is on a RP5470 running HP-UX 11.11. After logging in I do a "co" to get to the console. After typing in the login and password for root I get the following error and it kick me back to the login.

No utmp entry. You must exec "login" from the lowest level "sh"

Any help would be greatly appreciated!

Tim

[Basheer_2 · Accepted solution]

Hi Tim

I used to get thisin K220 box, I dont recall what exactly i did.

this is from the great ITRC forum
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1104160

[Ivan Krastev]

One other common reason for this error is corrupted /etc/utmp and /etc/utmpx files. To fix it these files have to be BOTH recreated.

regards,
ivan

[Aussan]

take a look at this doc KBRC00013729

http://www2.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c01016022-4

if the link does not work, here is the info

PROBLEM

When trying to login to the system (on a terminal, the console, by
telnet, ...)
or when running "exec login" at the shell prompt I get the following message :

No utmp entry. You must exec "login" from the lowest level "sh".

Why does this error occur?

RESOLUTION

There are many reasons causing login to display this message. Below is a list
of the most common solutions :

* If this happens when you run "exec login" at the shell prompt you are probably in a sub-shell of your login shell. Example : # ps -f UID PID PPID C STIME TTY TIME COMMAND root 27137 27136 0 11:12:37 ttyp2 0:00 -sh <== login shell root 294 274 1 11:38:01 ttyp2 0:00 ps -f root 274 27137 9 11:37:52 ttyp2 0:00 sh <== current sub-shell # exec login login: root Password: No utmp entry. You must exec "login" from the lowest level "sh". In such a case you must exit from the sub-shell before you can run the command.
* One other common cause for this error is corruption in the /etc/utmp and/or /etc/utmpx (this one is often overseen) files. To eliminate any corruption these files have to be BOTH recreated. The best way to do this is to insert two lines into the /etc/inittab file as follows, then save /etc/inittab and reboot : init:3:initdefault: utm1::sysinit:> /etc/utmp # clear current login accounting files utm2::sysinit:> /etc/utmpx # clear current login accounting files Note : simply rebooting the system does not recreate the files as might be stated in some documentation, it is mandatory to wipe them out using the above inittab entries. It is also strongly recommended to install the following patches to prevent new corruption of the files : 11.0 patches : PHCO_29633 : libc cumulative patch PHCO_22664 : cumulative patch for /sbin/init PHNE_26096 : telnet kernel and telnetd(1M) patch PHCO_19701 : tsm(1) cumulative patch PHSS_20864 : CDE Runtime MAR2000 Periodic Patch 11.11 patches : PHCO_30030 : libc cumulative patch PHNE_28841 : telnet kernel, telnetd(1M), telnet(1) patch Note: These patches, as with any patches, may be superseded. Please check for the latest patches at the IT Resource Center (ITRC) at the following web site:

http://www.itrc.hp.com
* Make sure your root filesystem has enough space. A full filesystem might prevent /etc/utmp(x) from growing in size.
* Verify that the /etc/inittab file does not contain lines with two character ids in the first field. For instance a line like this one : p1:234:respawn:/usr/sbin/getty -h tty0p1 9600 needs to be changed to : ttp1:234:respawn:/usr/sbin/getty -h tty0p1 9600 After such a modification run init q to activate the change.

[Timothy P. Jackson]

Thanks everyone for their responses!!

I will clear out the utmp, and the associated files then reboot.

Thanks again!

Tim