|
List Info
Thread: Warning: Current core (or exorcist perhaps) produces duplicate repligard table entries
|
|
| Warning: Current core (or exorcist
perhaps) produces duplicate repligard
table entries |

|
2006-03-02 12:52:57 |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi,
for all of you who are using the latest Midgard tech on
productive servers: Be
aware that some part of the package produces duplicate
entries in the repligard
table.
Midgard Core DOES NOT handle this case correctly. Versions
before something
around 1.7.3 do just ignore the case leading to the actual
duplicates (which
breaks MidCOM owner privileges and other stuff done directly
after create()),
while newer versions prevent object creation in such a case
(detected by an
UNIQUE INDEX, this is almost as bad as it prevents object
creation for obscure
reasons to users).
I am encountering this problem on a regular basis on at
least one live site, but
have not yet found a way to reproduce.
I thus recommend against using this piece of software until
this situation is
resolved.
Specifically I officially remove any recommendation to
deploy and MidCOM 2.5 on
live sites until this is resolved. I cannot support any of
these combinations
until either the root of this problem is solved or at least
the core gets more
error-resilent in this matter. This includes any bug
reporting where either
MidCOM 2.5 or Midgard 1.7 is concerned. Such bugs will not
be rejected if any
hint is there that points me to the conclusion that the
mentioned bug may be the
root of this problem.
Details can be found at the Midgard Bug Report #118 [1]
which also contains a
script [2] usable to detect and remove these duplicates in
case you absolutely
have to deploy 2.5 somewhere.
[1]
http://midgard.tigris.org/issues/show_bug.cgi?id=118
[2] http://midgard.tigris.org/nonav/issues
/showattachment.cgi/7/clear_guids.sh
Live long and Prosper!
Torben Nehmer
- --
Torben Nehmer, Guenzburg, Bavaria, Germany
http://www.nathan-syn
tronics.de, mailto:torben nehmer.net
PGP Public Key: https://www.li
nk-m.de/pgp/t.nehmer.asc
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFEBuqpJPh4Kn6d5FYRAtKaAJ9O8eRtNvGG7ZYzAjIdpvKOHI3c5QCg
/0Nb
n6wZfy1bcKv+wHtvPTH74PY=
=klNd
-----END PGP SIGNATURE-----
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe midgard-project.org
For additional commands, e-mail: dev-help midgard-project.org
|
|
| Warning: Current core (or exorcist
perhaps) produces duplicate repligard
table entries |

|
2006-03-02 12:58:38 |
Torben Nehmer <torben nehmer.net> wrote:
> while newer versions prevent object creation in such a
case (detected by an
> UNIQUE INDEX, this is almost as bad as it prevents
object creation for obscure
> reasons to users).
Can you copy here some url which points to documentaion
which says that object
and it's record is not created in such case?
Piotras
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe midgard-project.org
For additional commands, e-mail: dev-help midgard-project.org
|
|
| Warning: Current core (or exorcist
perhaps) produces duplicate repligard
table entries |

|
2006-03-02 13:37:54 |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi,
- --Piotras wrote on 2006-03-02 13:58:
>> while newer versions prevent object creation in
such a case (detected by an
>> UNIQUE INDEX, this is almost as bad as it prevents
object creation for obscure
>> reasons to users).
>
> Can you copy here some url which points to documentaion
which says that object
> and it's record is not created in such case?
You have been advertising just this as a solution in IRC.
Live long and Prosper!
Torben Nehmer
- --
Torben Nehmer, Guenzburg, Bavaria, Germany
http://www.nathan-syn
tronics.de, mailto:torben nehmer.net
PGP Public Key: https://www.li
nk-m.de/pgp/t.nehmer.asc
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFEBvUyJPh4Kn6d5FYRAik0AJ9vi7X/KH1iD2FCqlRwahzBed6o5QCf
Vwkw
nS14VvMRXPeUQEmNDoe+6tg=
=GKPG
-----END PGP SIGNATURE-----
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe midgard-project.org
For additional commands, e-mail: dev-help midgard-project.org
|
|
| Warning: Current core (or exorcist
perhaps) produces duplicate repligard
table entries |

|
2006-03-02 13:41:43 |
Torben Nehmer <torben nehmer.net> wrote:
> >> while newer versions prevent object creation
in such a case (detected by an
> >> UNIQUE INDEX, this is almost as bad as it
prevents object creation for obscure
> >> reasons to users).
> >
> > Can you copy here some url which points to
documentaion which says that object
> > and it's record is not created in such case?
>
> You have been advertising just this as a solution in
IRC.
Yes, I said about repligard's record not being created.
Because it doesn't matter what guid should be assigned for
object in such case.
If some guids for realm , id pair already exists then its
automagically assigned itself
to newly created object's record. Thus , no action is
required here as long as MySQL
prevents from duplicated records being created.
Piotras
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe midgard-project.org
For additional commands, e-mail: dev-help midgard-project.org
|
|
| Warning: Current core (or exorcist
perhaps) produces duplicate repligard
table entries |

|
2006-03-02 14:02:28 |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi,
- --Piotras wrote on 2006-03-02 14:41:
>>>> while newer versions prevent object
creation in such a case (detected
>>>> by an UNIQUE INDEX, this is almost as bad
as it prevents object
>>>> creation for obscure reasons to users).
>>>
>>> Can you copy here some url which points to
documentaion which says that
>>> object and it's record is not created in such
case?
>>
>> You have been advertising just this as a solution
in IRC.
>
> Yes, I said about repligard's record not being
created. Because it doesn't
> matter what guid should be assigned for object in such
case. If some guids
> for realm , id pair already exists then its
automagically assigned itself to
> newly created object's record. Thus , no action is
required here as long as
> MySQL prevents from duplicated records being created.
Great. You're reusing GUIDs possibly already in use in
other places. Welcome
Data Chaos. Go on, continue like this, and you'll have me
switch to another CMS
in not time.
Live long and Prosper!
Torben Nehmer
- --
Torben Nehmer, Guenzburg, Bavaria, Germany
http://www.nathan-syn
tronics.de, mailto:torben nehmer.net
PGP Public Key: https://www.li
nk-m.de/pgp/t.nehmer.asc
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFEBvr0JPh4Kn6d5FYRAr/tAKDjP70ORvLK1l2siNVzLWxb6kGJcACg
qq51
aQGQSt2HiSbqlnL0NDZc56w=
=cAcd
-----END PGP SIGNATURE-----
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe midgard-project.org
For additional commands, e-mail: dev-help midgard-project.org
|
|
[1-5]
|
|
|
about | contact Other archives ( Real Estate discussion Medical topics )
|