-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

steven.mcelwee@duke.edu writes:

> We are running PGP 6.5.8 on an Aix 5.3 server. About once per
>week, 15 files or so (each is 4 - 8 K in size) are encrypted
>concurrently and one will bomb. Our public keyring is tiny with only
>11 keys.
> Anyone have any ideas about this? keyring contention or some
>other bug, perhaps?

Do we assume these encryptions are being done in separate processes
which independently invoke PGP?

PGP 6.5.8 has a strange habit, where for every use it makes a
copy of the keyring, and uses that copy. It names these copies
"pubring-bak-n.pkr", where the "n" (as in "-n-" is a number between
1 and 9. It never keeps more than 9 of these, so it is deleting
some of them.

One guess is that maybe it is deleting one that is still in use by
another process.

Is there any reason not to use GPG? My guess is that it would be
more reliable for what you are doing.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)

iD8DBQFG0OnAvmGe70vHPUMRAp8lAKDb4ZA0SabrpOAgZkIsLe h60k0CPACfS+/C
KipdYzAcyp/Ct90hwXVASFU=
=AXB4
-----END PGP SIGNATURE-----

Yes, your assumption is correct that the encryptions are being run in
separate processes. I always wondered about the copies of the public
keyring. This is very useful to know.

We are interested in GPG and would like to know if the key format is
different than that in use with PGP 6.5.8.

thanks again,
Steven

"Neil W Rickert" wrote in message
news:ZQ5Ai.30479$RX.5996@newssvr11.news.prodigy.ne t...
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> steven.mcelwee@duke.edu writes:
>
>> We are running PGP 6.5.8 on an Aix 5.3 server. About once per
>>week, 15 files or so (each is 4 - 8 K in size) are encrypted
>>concurrently and one will bomb. Our public keyring is tiny with only
>>11 keys.
>> Anyone have any ideas about this? keyring contention or some
>>other bug, perhaps?
>
> Do we assume these encryptions are being done in separate processes
> which independently invoke PGP?
>
> PGP 6.5.8 has a strange habit, where for every use it makes a
> copy of the keyring, and uses that copy. It names these copies
> "pubring-bak-n.pkr", where the "n" (as in "-n-" is a number between
> 1 and 9. It never keeps more than 9 of these, so it is deleting
> some of them.
>
> One guess is that maybe it is deleting one that is still in use by
> another process.
>
> Is there any reason not to use GPG? My guess is that it would be
> more reliable for what you are doing.
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.2 (GNU/Linux)
>
> iD8DBQFG0OnAvmGe70vHPUMRAp8lAKDb4ZA0SabrpOAgZkIsLe h60k0CPACfS+/C
> KipdYzAcyp/Ct90hwXVASFU=
> =AXB4
> -----END PGP SIGNATURE-----
>

On Sun, 26 Aug 2007 02:47:21 GMT, Neil W Rickert
wrote:


>PGP 6.5.8 has a strange habit, where for every use it makes a
>copy of the keyring, and uses that copy. It names these copies
>"pubring-bak-n.pkr", where the "n" (as in "-n-" is a number between
>1 and 9. It never keeps more than 9 of these, so it is deleting
>some of them.
>

This is interesting. We have 4 copies of the pubring-bak-[1-4].pkr
file. Should we not have 9 copies?

>One guess is that maybe it is deleting one that is still in use by
>another process.
>
>Is there any reason not to use GPG? My guess is that it would be
>more reliable for what you are doing.

Are the public/private keys used in pgp 6.5.8 compatible with GPG? I
would like to avoid having to generate, distribute and acquire new
keys if possible.

Sincerely,
Steven McElwee

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

steven.mcelwee@duke.edu writes:

>Are the public/private keys used in pgp 6.5.8 compatible with GPG? I
>would like to avoid having to generate, distribute and acquire new
>keys if possible.

Yes, they are. However, if you are using a traditional RSA key,
then you will need the "idea" plugin for gpg.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)

iD8DBQFG16D6vmGe70vHPUMRAgD0AJwNmz3xrUf0UCIJ+kcm2j dLsZ/OUACgnFgO
SpcLQMpXFhVWLpvwkxWKqmw=
=1K9C
-----END PGP SIGNATURE-----