headers
Harihar Krishnan | 1 Sep 2004 13:09
Picon

SUMMARY: ASU permissions and Unix folders...


I got the solution for this tricky problem thanks to Dan Goetzman.

Harihar K
=======================================================
-----Original Message-----
From: Dan Goetzman [mailto:dan_goetzman <at> bmc.com]
Sent: Tuesday, August 31, 2004 9:27 PM
To: Harihar Krishnan
Subject: Re: ASU permissions and Unix folders...

Harihar,

I almost forgot to reply to this...

Looks like chacl is found at: /usr/sbin/chacl on my system.

copy/paste the following from my notes;

Reset a objects ACL using Unix command line
If the ACL entries are removed from a object, or set to certain values, 
then you will no longer be able to use the "net perm" command to change 
the ACL on a given object. To fix this use the ASU command "chacl" to 
either add back the ACE "Everyone=FullControl" or change the existing 
ACE for "Everyone" back to "FullControl".

* Add a ACE if there is not any existing ACE for "Everyone"

chacl -G rwaxdgsGSDcpoY:S-1-1-0 /bmc/export/...
(Continue reading)

Permalink | Reply | 0 comments |
headers
Ibrahim_Altahan | 1 Sep 2004 13:26
Picon

Printer question


Good morning all,
We are running Tru64 5.0A, and would like restrict a printer to specific
number of users.  Any idea how I can do this?  Thank you in advance.

Thank you,
Berhoom.
Permalink | Reply | 0 comments |
headers
Alaric Turner | 1 Sep 2004 17:01

Vrestore query

Managers,

DS20, Tru64 4.0G PK
I am doing a restore from tape & am getting an error from Vrestore

vrestore: can't get path for dir inode <23176>, file <file name> 

This is only happening on certain directories, & I think certain files.
The backup didn't report any problems on the files when backed up so I
am wondering what is going on?

Also are the files toast?

Many Thanks for any suggestions
Alaric
Permalink | Reply | 0 comments |
headers
James Hammett | 1 Sep 2004 17:53
Picon

Root can't change users NIS passwd

On my NIS master, not using enhanced security.  Using my root 
account, I am unable to change a NIS users account.

When I run yppasswd, it first asks for the users password. 
Unfortunately the user is brand new and has a "no login" for his 
password hash.  Afterwards the program responds "couldn't change 
passwd".  (I've also tried using the NIS master server's root 
password as the original password).

I'm sure there is relatively simple solution, but I can't seem to find it.

thanks,
James
--

-- 

--------------------------------------------------------------------------
James Hammett
Users Services / Server and Lab Administration (SLAM)
Information Technology Services ( (ITS)                          CPE 4.442
Chemical Engineering Unix Support                                   471-9701
----------------------------------------------------------------------------
        An injustice anywhere is a threat to justice everywhere - MLK jr.
----------------------------------------------------------------------------
Permalink | Reply | 0 comments |
headers
Zoong Pham | 2 Sep 2004 00:39

/etc/svc.conf and /etc/nsswitch.conf

In my server running 5.1B (with PK3), I have both /etc/svc.conf and 
/etc/nsswitch.conf
And there are man pages for nsswitch.conf(5) and svc.conf(4).

However, I install 5.1B-1 on another server today
It has only /etc/svc.conf and manpage of svc.conf(4).
Strange! There are no /etc/nsswitch.conf nor manpage for nsswitch.conf(5).
I then applied PK3 that installed only 1 patch (Patch 01139.00 - 
Patch resolves a symlink problem) out of 341 available 
patches. I guess others are already applied as part of 5.1B-1.

So my question is why /etc/nsswitch.conf and manpage for nsswitch.conf(5)
disappear in 5.1B-1?

I thought sometime in the past, one of the patch kits for version 5.X (prior
to 5.1B) added nsswitch.conf and left svc.conf for backward compatibility.

Here are the outputs from the 5.1B-1 server:
  # sizer -v
  Compaq Tru64 UNIX V5.1B (Rev. 2650); Tue Aug 31 15:52:04 EST 2004
  # uname -a
  OSF1 alpha3.mercy.com.au V5.1 2650 alpha

Can someone clarify this for me?

Thanks,
Zoong
Permalink | Reply | 0 comments |
headers
Hazem Tolba | 2 Sep 2004 09:52

DEC Server 90TL

Hi Managers,

Could any one help us on how to install the image file for the decserver
90TL on Tru64 Unix alpha system?
We have one customer how have such servers and they were booting from an
old alpha 2100 running Tru64 unix v5.1. Now the customer has bought a
new alpha ES45 running Tru64 unix v5.1B and he needs to make the 90TL
servers to boot (or download) their image form the new alpha so that he
can completely get rid of the old alpha.
We have tried to copy all the necessary files (from our point of view)
from the old to the new alpha (inetd.conf, bootptab, hosts, mneng1.sys)
and almost having the same configuration on the new alpha, but still the
90TL servers can't load their image from the new alpha.

Could anyone help us how to solve this problem?
Thanks all and BR.
Permalink | Reply | 2 comments |
headers
Rich Glazier | 2 Sep 2004 19:01
Picon
Favicon

setting mtu per route

Is there any way to set an MTU size per a static
route, like there is in other OSes?  I have a FDDI
card that has a default 4500 PMTU size.  I would like
a specific route from that interface to transmit at no
larger than 1500.  Is that doable?

Thanks 

	
		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail
Permalink | Reply | 1 comment |
headers
Rich Glazier | 2 Sep 2004 20:53
Picon
Favicon

UPDATE: setting mtu per route

I set the route to be locked at 1500.

mars# netstat -rn | grep 129.250.150.121
129.250.150.121  10.1.101.251       UGHD        2     
763  ee0

mars# route delete 129.250.150.121 10.1.101.251
delete host 129.250.150.121: gateway 10.1.101.251

mars# route add -host 129.250.150.121 10.1.101.251
-lock -mtu 1500
add host 129.250.150.121: gateway 10.1.101.251

mars# netstat -rnv | grep 129.250.150.121
129.250.150.121  10.1.101.251       UGHS        1     
512  ee0         1500*     0      0      0      0     
0       0       0      0

Should a route add take effect immediately, or would I
need to restart networking?

	
		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail
Permalink | Reply | 0 comments |
headers
Marco Lolli | 3 Sep 2004 11:44
Picon

Summary: ftp server on Alpha TRU64 5.1B


	I had received a few replies from

	Trevor Osatchuk
	
	Patrick Miguel

	and Thomas Blinn

	thanks to everyone

	First of all they say to check all the network
	configuration both hardware and software, but
	all it is ok, Thomas pointed out that could be
	a gftp bug, or a "misunderstanding" between the 
	client and the server indeed also gftp distribuited 		
	with Mandrake 10.0 experienced the same behaviour
	with ftpd server on TRU64 5.1B.

	Thomas says also that HP could test ftp server
	within his environment, I tried another 
	gui ftp client as Kbear and all it is ok.

	Marco Lolli

	Bologna Astronomical Observatory

	My Original Question:

ftp server on Alpha TRU64  5.1B
(Continue reading)

Permalink | Reply | 0 comments |
headers
Sdrolias Aristotelis | 3 Sep 2004 17:19
Picon

Memory delays in GS160 OS:Tru64 5.1 PK3.

Hello all, 

System: GS160 OS:Tru64 5.1 PK3.

I have some questions on memory delays and how they can impact performance of the system.
Any help would be valueable.

---------------------------------------------------
Memory Interleaving 
---------------------------------------------------

I got this info from HP:
"Mixing the memory sizes limits the interleaving capability and the potential bandwidth of the system.
Better system performance can be achieved with a larger interleaving factor. 
For better memory bandwidth, spread two, four, or eight arrays (a power of two) evenly across the four
memory modules."

1) If I understand correctly the memory chips used should be of the same size  and their number should be a
power of 2 in one particular QBB. If not what is the optimum placement and size selection?
2) If I use memory chips of different size on the same QBB this might potentially reduce the interleaving
factor as stated in  the answer I got from HP. This means that the pipelining that "memory interleaving" is
trying to establish is not functioning as it should be.
Are there any other possible impacts? (e.g redirection of memory request to an other RAD?) 
How this can impact the performance of the system and in what level?

Kind regards & have a nice weekend, 
---------------------------------------------------
A. Sdrolias
email: asdrolias <at> cosmote.gr
(Continue reading)

Permalink | Reply | 0 comments |

Gmane