Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Chaskiel M Grundman]

--On Friday, April 15, 2005 10:42:53 PM -0400 chas williams - CONTRACTOR 
<[EMAIL PROTECTED]> wrote:

In message <[EMAIL PROTECTED]>,Chaskiel M Grundman writes:
--==E0A3C7D627E180487321==
I'm not claiming there isn't a problem. I'm claiming that whatever
problem=20 exists is not preallocs that are not being discarded because
of the=20 truncate optimization.
Well, what I said was correct, but you were definitely on the right track. 
Try this:
(I'll put 'we should try to use dentry_open/filp_close instead of mucking 
about on our own' on my ever-growing todo list of afs fixes)

Index: LINUX/osi_file.c
===
RCS file: /cvs/openafs/src/afs/LINUX/osi_file.c,v
retrieving revision 1.23
diff -u -d -r1.23 osi_file.c
--- LINUX/osi_file.c11 Mar 2005 04:35:42 -  1.23
+++ LINUX/osi_file.c16 Apr 2005 04:12:30 -
@@ -65,6 +65,7 @@
filp->f_mapping = tip->i_mapping;
#endif
#if defined(AFS_LINUX24_ENV)
+filp->f_mode = FMODE_READ|FMODE_WRITE;
filp->f_op = fops_get(tip->i_fop);
#else
filp->f_op = tip->i_op->default_file_ops;



p7seCBK5eNcIY.p7s
Description: S/MIME cryptographic signature

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[chas williams - CONTRACTOR]

In message <[EMAIL PROTECTED]>,Chaskiel M Grundman writes:
>--==E0A3C7D627E180487321==
>I'm not claiming there isn't a problem. I'm claiming that whatever problem=20
>exists is not preallocs that are not being discarded because of the=20
>truncate optimization.

dont know.  the ext2 behavior's is rather strange then.  looking at my
cache, i see things like:

 8 -rw---  1 root root  2048 Apr 15 22:27 V6123
 8 -rw---  1 root root  2048 Apr 15 22:27 V6125
 8 -rw---  1 root root  2048 Apr 15 22:27 V6126
 8 -rw---  1 root root  2048 Apr 15 22:27 V6129
 8 -rw---  1 root root  2048 Apr 15 22:27 V6131
 8 -rw---  1 root root  4096 Apr 15 22:25 V4506
 8 -rw---  1 root root  4096 Apr 15 22:25 V4967
 8 -rw---  1 root root  4096 Apr 15 22:25 V5472
 8 -rw---  1 root root  4096 Apr 15 22:25 V5473

8 blocks to store 2048 bytes?  normally i would just say its the
filesystem trying to helpful, but creating a similar file with dd, shows

#  dd if=/dev/zero bs=2048 of=/cache/dd.out count=1
1+0 records in
1+0 records out
# ls -ls /cache/dd.out 
 2 -rw-r--r--  1 root root 2048 Apr 15 22:33 /cache/dd.out

>afs holds one file open: the CacheItems file (on bsd's, it holds two files=20
>open: the CacheItems file and the VolumeItems file). That's it. other calls =

you learn something new every day.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Chaskiel M Grundman]

--On Friday, April 15, 2005 07:55:37 PM -0400 chas williams - CONTRACTOR 
<[EMAIL PROTECTED]> wrote:

In message <[EMAIL PROTECTED]>,Chaskiel M
Grundman writes:
afs does not hold cache files open. Most calls to afs_CFileTruncate
are=20 immediately followed by afs_CFileClose (which is osi_UFSClose,
which calls=20 release_file).
that's not my impression.  /cache is an ext2 filesystem:
[...]
I'm not claiming there isn't a problem. I'm claiming that whatever problem 
exists is not preallocs that are not being discarded because of the 
truncate optimization.

i didnt include getcacheparms but it only thinks the cache increased by 1k
but in reality it went up by 8k.  isnt there a list of open/active cache
files (dcache slots?)?
afs holds one file open: the CacheItems file (on bsd's, it holds two files 
open: the CacheItems file and the VolumeItems file). That's it. other calls 
to afs_CFileOpen/afs_CFileClose are balanced

% xstat_cm_test localhost -once 0 | egrep 'osi_UFSOpen|osi_Close'
   344299 osi_UFSOpen
   344298 osi_Close


p7sPtrMlNIeEj.p7s
Description: S/MIME cryptographic signature

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[chas williams - CONTRACTOR]

In message <[EMAIL PROTECTED]>,Chaskiel M Grundman writes:
>afs does not hold cache files open. Most calls to afs_CFileTruncate are=20
>immediately followed by afs_CFileClose (which is osi_UFSClose, which calls=20
>release_file).

that's not my impression.  /cache is an ext2 filesystem:

relax.7% ls -l short.bonnie.sh
-rwxr-xr-x  1 chas users 55 Aug 18  2002 tmp/short.bonnie.sh
relax.8% fs flush short.bonnie.sh
relax.9% df -k /cache
Filesystem   1K-blocks  Used Available Use% Mounted on
/dev/sdb1   482214427736 29579  94% /cache
relax.10% cp short.bonnie.sh /dev/null
relax.11% df -k /cache
Filesystem   1K-blocks  Used Available Use% Mounted on
/dev/sdb1   482214427744 29571  94% /cache

i didnt include getcacheparms but it only thinks the cache increased by 1k
but in reality it went up by 8k.  isnt there a list of open/active cache
files (dcache slots?)?  the bulk of cache files are closed but if you had
a fullish cache and opened a bunch of small files you could overextend.

ext2 works fine as a cache if i limit it to 1/3 to 1/2 the size of the
cache partition.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Chaskiel M Grundman]

--On Friday, April 15, 2005 18:42:14 -0400 chas williams - CONTRACTOR 
<[EMAIL PROTECTED]> wrote:

In message <[EMAIL PROTECTED]>,Chaskiel M
Grundman  writes:
It should not be necessary to call ext2_truncate to free the
preallocated  blocks - ext2_release_file will do it.
but you seem to get these preallocated blocks if you just open a file
for writing.  so if your disk cache consists lots of tiny files you
wind up with oversubscribing the cache partition.
I don't see that. I'm looking at 2.6.11.5, and the only way for 
preallocation to happen is for someone to call ext2_get_block. neither 
ext2_read_inode or generic_file_open (which ext2 uses) result in any of the 
aops being called.

afs does not hold cache files open. Most calls to afs_CFileTruncate are 
immediately followed by afs_CFileClose (which is osi_UFSClose, which calls 
release_file).




p7sbsPYeMzWKN.p7s
Description: S/MIME cryptographic signature

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[chas williams - CONTRACTOR]

In message <[EMAIL PROTECTED]>,Chaskiel M Grundman
 writes:
>It should not be necessary to call ext2_truncate to free the preallocated 
>blocks - ext2_release_file will do it.

but you seem to get these preallocated blocks if you just open a file
for writing.  so if your disk cache consists lots of tiny files you
wind up with oversubscribing the cache partition.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Chaskiel M Grundman]

--On Friday, April 15, 2005 17:59:14 -0400 chas williams - CONTRACTOR 
<[EMAIL PROTECTED]> wrote:

so ext2_truncate() isnt called on cache file open to get rid of
the preallocations.   or so i think.
It should not be necessary to call ext2_truncate to free the preallocated 
blocks - ext2_release_file will do it.

p7shf2HO5VT7Q.p7s
Description: S/MIME cryptographic signature

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[chas williams - CONTRACTOR]

In message <[EMAIL PROTECTED]>,Kris Van Hees writes:
>I wonder if ext2 itself might defer truncates as well in newer kernels.  I'll
>have a look at that.

it looks like we are getting bitten by preallocation in the ext2 
filesystem combined with an optimization in inode_setattr().  2.6's
inode_setattr has the following test:

if (attr->ia_size != i_size_read(inode)) {
error = vmtruncate(inode, attr->ia_size);
if (error || (ia_valid == ATTR_SIZE))
goto out;
} else {

so ext2_truncate() isnt called on cache file open to get rid of 
the preallocations.   or so i think.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] deleting volumes

[Jeffrey Altman]

Ron Croonenberg wrote:

> Hello,
> 
> I used the "AFS server manager (1.3.77)" to delete some volumes, but after I
> deleted them I still see their mountpoints (users home directory) but when I 
> do
> an "ls /home/user" I get :
> ls: user: No such device
> 
> How do I fix that ?
> 
> thanks,
> 
> Ron

Symlinks and mountpoints which pointed at non-existent destinations were
not reported as directories in OpenAFS for Windows versions prior to
1.3.80.

Jeffrey Altman

P.S. - I'm still waiting to receive a bug report from you on your last
complaint.




smime.p7s
Description: S/MIME Cryptographic Signature

Re[3]: [OpenAFS] deleting volumes

[Ron Croonenberg]

Well the read/write copy is in /afs/.mycell

/home is a symbolic link to /afs/mycell



>On Fri, 15 Apr 2005, Ron Croonenberg wrote:
>
>> thanks for the quick response, I appreciate it
>>
>> when I do a :
>>
>> fs rmm /home/user I get:
>> fs: '/home/user' is not a mount point.
>
>
>unless you mounted afs on /home, that's necessarily true
>
>> when I do :
>> fs rmm user
>> fs: You can not change a backup or readonly volume
>
>yup. where's the read-write copy of the volume? Probably something like
>/afs/.yourcell/home ?
>___
>OpenAFS-info mailing list
>OpenAFS-info@openafs.org
>https://lists.openafs.org/mailman/listinfo/openafs-info

=
1879:
 Thomas Edison gets an idea, and his brother Timmy says,
 "Hey, what's that thing over your head?
=
 Ron Croonenberg   | Phone: 1 765 658 4761
 Technology Coordinator| Fax:   1 765 658 4732
   |
 Department of ComputerScience | e-mail : [EMAIL PROTECTED]
 DePauw University |
 Julian Science & Math Center  |
 602 South College Ave.|
 Greencastle, IN  46135|
=
 http://www.depauw.edu/acad/computer/RonCroonenberg.asp
=
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re[2]: [OpenAFS] deleting volumes

[Derrick J Brashear]

On Fri, 15 Apr 2005, Ron Croonenberg wrote:
thanks for the quick response, I appreciate it
when I do a :
fs rmm /home/user I get:
fs: '/home/user' is not a mount point.

unless you mounted afs on /home, that's necessarily true
when I do :
fs rmm user
fs: You can not change a backup or readonly volume
yup. where's the read-write copy of the volume? Probably something like 
/afs/.yourcell/home ?
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] deleting volumes

[Derrick J Brashear]

On Fri, 15 Apr 2005, Ron Croonenberg wrote:
Hello,
I used the "AFS server manager (1.3.77)" to delete some volumes, but after I
deleted them I still see their mountpoints (users home directory) but when I do
an "ls /home/user" I get :
ls: user: No such device
the opposite of fs mkmount is fs rmmount
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] deleting volumes

[Russ Allbery]

Ron Croonenberg <[EMAIL PROTECTED]> writes:

> thanks for the quick response, I appreciate it
> when I do a :

> fs rmm /home/user I get:
> fs: '/home/user' is not a mount point.

Is /home/user a symlink into AFS?

> when I do :
> fs rmm user
> fs: You can not change a backup or readonly volume

You may need to explicitly go to the read-write path for your cell (by
putting a period before your cell name; /afs/.ir.stanford.edu instead of
/afs/ir.stanford.edu, for instance) in order to remove the mount point if
the parent volume is replicated.

If that's the case, you'll then want to vos release the parent volume
after you remove the mount point.

-- 
Russ Allbery ([EMAIL PROTECTED]) 
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re[2]: [OpenAFS] deleting volumes

[Ron Croonenberg]

thanks for the quick response, I appreciate it

when I do a :

fs rmm /home/user I get:
fs: '/home/user' is not a mount point.

when I do :
fs rmm user
fs: You can not change a backup or readonly volume


thanks,

Ron


>Just as creating a volume doesn't automagically mount it, deleting a
>volume doesn't automatically remove the mount point.
>
>On a Unix box you'd do "fs rmm mountpointname.
>
>
>
>On 4/15/05, Ron Croonenberg <[EMAIL PROTECTED]> wrote:
>> Hello,
>>
>> I used the "AFS server manager (1.3.77)" to delete some volumes, but afte
>r I
>> deleted them I still see their mountpoints (users home directory) but whe
>n I do
>> an "ls /home/user" I get :
>> ls: user: No such device
>>
>> How do I fix that ?
>>
>> thanks,
>>
>> Ron
>>
>> 
>=
>
>> 1879:
>>  Thomas Edison gets an idea, and his brother Timmy says,
>>  "Hey, what's that thing over your head?
>> 
>=
>
>>  Ron Croonenberg   | Phone: 1 765 658 4761
>>  Technology Coordinator| Fax:   1 765 658 4732
>>|
>>  Department of ComputerScience | e-mail : [EMAIL PROTECTED]
>>  DePauw University |
>>  Julian Science & Math Center  |
>>  602 South College Ave.|
>>  Greencastle, IN  46135|
>> 
>=
>
>>  http://www.depauw.edu/acad/computer/RonCroonenberg.asp
>> 
>=
>
>> ___
>> OpenAFS-info mailing list
>> OpenAFS-info@openafs.org
>> https://lists.openafs.org/mailman/listinfo/openafs-info
>>
>___
>OpenAFS-info mailing list
>OpenAFS-info@openafs.org
>https://lists.openafs.org/mailman/listinfo/openafs-info

=
1879:
 Thomas Edison gets an idea, and his brother Timmy says,
 "Hey, what's that thing over your head?
=
 Ron Croonenberg   | Phone: 1 765 658 4761
 Technology Coordinator| Fax:   1 765 658 4732
   |
 Department of ComputerScience | e-mail : [EMAIL PROTECTED]
 DePauw University |
 Julian Science & Math Center  |
 602 South College Ave.|
 Greencastle, IN  46135|
=
 http://www.depauw.edu/acad/computer/RonCroonenberg.asp
=
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] deleting volumes

[Esther Filderman]

Just as creating a volume doesn't automagically mount it, deleting a
volume doesn't automatically remove the mount point.

On a Unix box you'd do "fs rmm mountpointname."  



On 4/15/05, Ron Croonenberg <[EMAIL PROTECTED]> wrote:
> Hello,
> 
> I used the "AFS server manager (1.3.77)" to delete some volumes, but after I
> deleted them I still see their mountpoints (users home directory) but when I 
> do
> an "ls /home/user" I get :
> ls: user: No such device
> 
> How do I fix that ?
> 
> thanks,
> 
> Ron
> 
> =
> 1879:
>  Thomas Edison gets an idea, and his brother Timmy says,
>  "Hey, what's that thing over your head?
> =
>  Ron Croonenberg   | Phone: 1 765 658 4761
>  Technology Coordinator| Fax:   1 765 658 4732
>|
>  Department of ComputerScience | e-mail : [EMAIL PROTECTED]
>  DePauw University |
>  Julian Science & Math Center  |
>  602 South College Ave.|
>  Greencastle, IN  46135|
> =
>  http://www.depauw.edu/acad/computer/RonCroonenberg.asp
> =
> ___
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
>
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

[OpenAFS] deleting volumes

[Ron Croonenberg]

Hello,

I used the "AFS server manager (1.3.77)" to delete some volumes, but after I
deleted them I still see their mountpoints (users home directory) but when I do
an "ls /home/user" I get :
ls: user: No such device

How do I fix that ?

thanks,

Ron

=
1879:
 Thomas Edison gets an idea, and his brother Timmy says,
 "Hey, what's that thing over your head?
=
 Ron Croonenberg   | Phone: 1 765 658 4761
 Technology Coordinator| Fax:   1 765 658 4732
   |
 Department of ComputerScience | e-mail : [EMAIL PROTECTED]
 DePauw University |
 Julian Science & Math Center  |
 602 South College Ave.|
 Greencastle, IN  46135|
=
 http://www.depauw.edu/acad/computer/RonCroonenberg.asp
=
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] 1.3.81 under AIX 5.1

[Steve Roseman]

Derrick J Brashear wrote:
On Fri, 15 Apr 2005, Hans-Gunther Borrmann wrote:
"make dest" terminates with an error message:

"terminates". The error is:
make: 1254-005 Ignored error code 1 from last command.

ignored.
afs_dynamic_auth and afs_dynamic_kerbauth are completely missing. The 
1.3.80
versions seem to work.

Will be fixed in the next version.
I changed 2 lines in Makefile.in to resolve both of these:
56c56
<   -${INSTALL} ${srcdir}/doc/LICENSE ${DEST}/LICENSE
---
>   -${INSTALL} ${srcdir}/src/LICENSE ${DEST}/LICENSE
558c558
<   libafsauthent shlibafsrpc shlibafsauthent libadmin
---
>   libafsauthent shlibafsrpc shlibafsauthent libadmin tsm41
Otherwise, AFS (client) has been working on my test systems.  Thanks to 
those who did all of the work.

My very minor "problem" is not being able to tell the difference between 
PAG and UID-based tokens.  "groups" no longer lists the special groups. 
 IBM resolved it by changing the output of AIX 5 "tokens" to indicate 
PAG or UID based tokens in the output.  I'd be willing to to do the same 
if I had a quick "here's what to look for".

Steve
-
Stephen G. Roseman
Lehigh University
[EMAIL PROTECTED]
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Problem with pam on debian with 1.3.81 kernel 2.6.11

[Derek Atkins]

Quoting "Douglas E. Engert" <[EMAIL PROTECTED]>:

> 
> Do you have the log on both sides?

No, only the client log.
 
> These was ambiguity with the early drafts as to how an OID was to be passed,
> and I have this additional mod to OpenSSH-3.8 which tries both ways.
> But the logit should have written out the Badly formed OID message
> to the syslog if this was the problem. Do you see it in the syslog?

Is this a change on the client or the server?  I see nothing in the server
syslogs about any kind of OID message, or indeed anything at all.

-derek

-- 
   Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
   Member, MIT Student Information Processing Board  (SIPB)
   URL: http://web.mit.edu/warlord/PP-ASEL-IA N1NWH
   [EMAIL PROTECTED]PGP key available

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Problem with pam on debian with 1.3.81 kernel 2.6.11

[Douglas E. Engert]

Derek Atkins wrote:
"Douglas E. Engert" <[EMAIL PROTECTED]> writes:

There is a gssapimitm.patch for OpenSSH-3.8 that will let it
do both if you set: "GSSAPIEnableMITMAttack yes". Its from March 2004.
So you can interoperate if you update the old server, or add this
patch as a conversion aid and live with the problem for a while.

Thanks for the pointer.  It turns out that FC3's openssh includes
this patch but does not build with it.  There's one bug in FC3's
version of the patch; I had to change one hunk to get it to apply.
In particular, I needed to add the ' sAcceptEnv,' in the following
two lines of one of the hunks:
-   sGssAuthentication, sGssCleanupCreds, sAcceptEnv,
+   sGssAuthentication, sGssCleanupCreds, sAcceptEnv, sGssEnableMITM,
With this change the patch applied and I now have FC3 RPMs.
Unfortuately it appears not to be working properly.  It's not
acquiring any tickets on the client and it's not proceeding with the
GSSAPI negotiation.  :(
debug1: Authentications that can continue: external-keyx,gssapi,password
debug3: start over, passed a different list external-keyx,gssapi,password
debug3: preferred 
gssapi-with-mic,gssapi,publickey,keyboard-interactive,passworddebug3: 
authmethod_lookup gssapi
debug3: remaining preferred: publickey,keyboard-interactive,password
debug3: authmethod_is_enabled gssapi
debug1: Next authentication method: gssapi
debug2: we sent a gssapi packet, wait for reply
debug1: Authentications that can continue: external-keyx,gssapi,password
debug2: we sent a gssapi packet, wait for reply
debug1: Authentications that can continue: external-keyx,gssapi,password
debug2: we did not send a packet, disable method
debug3: authmethod_lookup password
debug3: remaining preferred: ,keyboard-interactive,password
debug3: authmethod_is_enabled password
debug1: Next authentication method: password
-derek
Do you have the log on both sides?
These was ambiguity with the early drafts as to how an OID was to be passed,
and I have this additional mod to OpenSSH-3.8 which tries both ways.
But the logit should have written out the Badly formed OID message
to the syslog if this was the problem. Do you see it in the syslog?

--- ,auth2-gss.cWed Mar  3 13:21:18 2004
+++ auth2-gss.c Fri Mar  5 14:38:25 2004
@@ -90,7 +90,22 @@
gss_test_oid_set_member(&ms, &oid, supported,
&present);
} else {
+#if 1
+   {
+extern char * client_version_string;
+   /* should only do for SecureCRT 4.0 */
+   if (client_version_string) {
+   logit("Badly formed OID received from %s using 
%s",
+   authctxt->user,
+   client_version_string);
+   }
+   oid.elements = doid;
+   oid.length = len;
+   gss_test_oid_set_member(&ms, &oid, supported,&present);
+   }
+#else
logit("Badly formed OID received");
+#endif
}
} while (mechs > 0 && !present);

--
 Douglas E. Engert  <[EMAIL PROTECTED]>
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[chas williams - CONTRACTOR]

In message <[EMAIL PROTECTED]>,Hans-Werner Paulse
n writes:
>And the cache partition is ext2.

another datapoint.  i was able to duplicate this problem.  however,
the cache partition had to be ext2.  i didnt see the same problem
using an ext3 cache.

SMP ([EMAIL PROTECTED]) 2.6.11.7, OpenAFS 1.3.81

any chance i could convince you to try an ext3 cache while i try to
see what's broken with ext2.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Kris Van Hees]

On Fri, Apr 15, 2005 at 12:02:39PM +0100, Dr A V Le Blanc wrote:
> On Thu 14 Apr 2005 at 10:46:47 -0400, Kris Van Hees <[EMAIL PROTECTED]> wrote:
> > I am almost willing to bet that the cache partition in this case is ext3 or
> > any other jfs.
> 
> No, the cache partition is, and always was, ext2.

Ah, thanks for that information.  That certainly changes the situation a lot,
and definitely indicates there is something else going wrong beyond the jfs
problem I found not too long ago.

I wonder if ext2 itself might defer truncates as well in newer kernels.  I'll
have a look at that.

Kris
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Problem with pam on debian with 1.3.81 kernel 2.6.11

[Derek Atkins]

"Douglas E. Engert" <[EMAIL PROTECTED]> writes:

> There is a gssapimitm.patch for OpenSSH-3.8 that will let it
> do both if you set: "GSSAPIEnableMITMAttack yes". Its from March 2004.
>
> So you can interoperate if you update the old server, or add this
> patch as a conversion aid and live with the problem for a while.

Thanks for the pointer.  It turns out that FC3's openssh includes
this patch but does not build with it.  There's one bug in FC3's
version of the patch; I had to change one hunk to get it to apply.

In particular, I needed to add the ' sAcceptEnv,' in the following
two lines of one of the hunks:

-   sGssAuthentication, sGssCleanupCreds, sAcceptEnv,
+   sGssAuthentication, sGssCleanupCreds, sAcceptEnv, sGssEnableMITM,

With this change the patch applied and I now have FC3 RPMs.
Unfortuately it appears not to be working properly.  It's not
acquiring any tickets on the client and it's not proceeding with the
GSSAPI negotiation.  :(

debug1: Authentications that can continue: external-keyx,gssapi,password
debug3: start over, passed a different list external-keyx,gssapi,password
debug3: preferred 
gssapi-with-mic,gssapi,publickey,keyboard-interactive,passworddebug3: 
authmethod_lookup gssapi
debug3: remaining preferred: publickey,keyboard-interactive,password
debug3: authmethod_is_enabled gssapi
debug1: Next authentication method: gssapi
debug2: we sent a gssapi packet, wait for reply
debug1: Authentications that can continue: external-keyx,gssapi,password
debug2: we sent a gssapi packet, wait for reply
debug1: Authentications that can continue: external-keyx,gssapi,password
debug2: we did not send a packet, disable method
debug3: authmethod_lookup password
debug3: remaining preferred: ,keyboard-interactive,password
debug3: authmethod_is_enabled password
debug1: Next authentication method: password

-derek

-- 
   Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
   Member, MIT Student Information Processing Board  (SIPB)
   URL: http://web.mit.edu/warlord/PP-ASEL-IA N1NWH
   [EMAIL PROTECTED]PGP key available
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] 1.3.81 under AIX 5.1

[Derrick J Brashear]

On Fri, 15 Apr 2005, Hans-Gunther Borrmann wrote:
"make dest" terminates with an error message:
"terminates". The error is:
make: 1254-005 Ignored error code 1 from last command.
ignored.
afs_dynamic_auth and afs_dynamic_kerbauth are completely missing. The 1.3.80
versions seem to work.
Will be fixed in the next version.
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

[OpenAFS] 1.3.81 under AIX 5.1

[Hans-Gunther Borrmann]

Hello,

I tested the 1.3.81 client under AIX 5.1,single processor, 32-Bit kernel. Here 
my results:

It is the first OpenAFS version which survives the first klog under AIX 5.1. 
Thanks to all who contributed. I did some stress tests. The client seems to 
work well.

"make dest" terminates with an error message:
/u/b/borrmann/sw/openafs/1.3.81/aix/openafs-1.3.81/src/pinstall/pinstall ./
doc/LICENSE /u/b/borrmann/sw/openafs/1.3.81/aix/openafs-1.3.81/rs_aix51/dest/
LICENSE
Can't open source file ``./doc/LICENSE'': No such file or directory
make: 1254-004 The error code from the last command is 1.
make: 1254-005 Ignored error code 1 from last command.

afs_dynamic_auth and afs_dynamic_kerbauth are completely missing. The 1.3.80 
versions seem to work.

Gunther Borrmann
-- 

Hans-Gunther Borrmann <[EMAIL PROTECTED]>
Rechenzentrum der Universitaet Freiburg
Hermann-Herder-Str. 10, D79104 FREIBURG
Tel.: +49 761/203-4652
Fax:  +49 761/203-4643

___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

Re: [OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Hans-Werner Paulsen]

On Fri, Apr 15, 2005 at 12:02:39PM +0100, Dr A V Le Blanc wrote:
>...
> rebooted the system and gone into afs and done a
> 
>  for i in `find . -type f -noleaf`;do cat $i >/dev/null;done
> 
> The cache partition quickly gets overfull and input/output errors
> appear.  This looks like a real problem.  I am now using the
> Debian packakes for 1.3.81, but previously I compiled openafs
> from source and had the same problem, so it does not appear to
> be specific to the Debian packages.
> 
> Is anyone else running 1.3.81 on 2.6.11 of some form, who is willing
> to try the same test?
> 

I have the same problem with my machine:
linux-2.6.11.5
openafs-1.3.81
gcc-3.3.2
glibc-2.3.2
This is a self-made system using unmodified sources.

The three commands "df", "du" and "fs getcache" are showing different
results (unit=1k)
df: 669132 of 1035692
fs: 432785 of 50 (I reduced the max size to avoid problems)
du: 2520660 (!!!)

And the cache partition is ext2.

-- 
Hans-Werner Paulsen [EMAIL PROTECTED]
MPI für Astrophysik Tel 089-3-2602
Karl-Schwarzschild-Str. 1   Fax 089-3-2235  
D-85741 Garching
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info

[OpenAFS] Re: Problem with openafs-1.3.81 on kernel 2.6.11.7

[Dr A V Le Blanc]

On Thu 14 Apr 2005 at 10:46:47 -0400, Kris Van Hees <[EMAIL PROTECTED]> wrote:
> I am almost willing to bet that the cache partition in this case is ext3 or
> any other jfs.

No, the cache partition is, and always was, ext2.

On Thu 14 Apr 2005 at 10:34:09 -0400, Chas Williams <[EMAIL PROTECTED]> wrote:
> well it would be a good idea to mkfs the filesystem before you start
> 2.6/1.3.81 again just to make sure there are not residual bits.  however,
> i would guess that looks like truncate's arent happening in a timely
> fashion.  i believe someone else on the list has seen this behavior as
> well and might have a patch to make truncate() synchronize.
> 
> just curious, is there any chance you could try a memcache?  you wouldnt
> need a very big one.

I have now done a mkfs on the cache partition and rebooted.  I've also
made certain that both /etc/openafs/cacheinfo and the /etc/openafs/afs.conf
contain explicit cache limits; both say 30, and the partition is
still a 500mb partition.  (Is this the right way to specify 300mb?
I can't find any use of the CACHE variable anywhere.)  I've then
rebooted the system and gone into afs and done a

 for i in `find . -type f -noleaf`;do cat $i >/dev/null;done

The cache partition quickly gets overfull and input/output errors
appear.  This looks like a real problem.  I am now using the
Debian packakes for 1.3.81, but previously I compiled openafs
from source and had the same problem, so it does not appear to
be specific to the Debian packages.

Is anyone else running 1.3.81 on 2.6.11 of some form, who is willing
to try the same test?

 -- Owen
 [EMAIL PROTECTED]
___
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info