I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer, *indicating* that the *command* *is* done."
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
But as these exceptions are marked as Resumable, shouldn't it be enough for me to do :- [ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck marianopeck@gmail.comwrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the
connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the
connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck marianopeck@gmail.comwrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the
connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the
connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
Well the VM crashes, but I can't be sure if its because of the resume or something else.
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck marianopeck@gmail.comwrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer,
*indicating* that the *command* *is* done."
This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if the
connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
On Wed, May 11, 2011 at 3:54 PM, John Toohey jt@parspro.com wrote:
Well the VM crashes, but I can't be sure if its because of the resume or something else.
wow.... we will need more data in order to help there. Which Pharo/Squeak image, which VM ? which OS ? which OpenDBX version ? which database ? which version of SqueakDBX/ConfigurationOf ?
do you have the dump or console output of the crash ?
thanks
Mariano
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
> I get this error a lot when my system is under load. According to the > pgsql docs, I should do this, "PQgetResult must be called repeatedly > until it returns a null pointer, >
*indicating* that the *command* *is* done." > > This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if
the connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
> How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 > and the Glorp/DBX build that was available when that was released. > > -- > ~JT > > > > _______________________________________________ > SqueakDBX mailing list > SqueakDBX@lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx > >
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
- Pharo-1.1.1--Latest update: #11414 - Posstgres 8.3.7 - 32 bit compiled OpenDBX from approx 12 months ago - Headless Linux server running on EC2 - VM is 15.2f
What I often get it the debugger firing with Glorp errors because the Glorp proxy does not instantiate an object associated with a foreign key. I can usually get out of those, but often in the morning, the VM itself has gone down. Attached is a PharoDebug from this morning.
On Wed, May 11, 2011 at 10:26, Mariano Martinez Peck marianopeck@gmail.comwrote:
On Wed, May 11, 2011 at 3:54 PM, John Toohey jt@parspro.com wrote:
Well the VM crashes, but I can't be sure if its because of the resume or something else.
wow.... we will need more data in order to help there. Which Pharo/Squeak image, which VM ? which OS ? which OpenDBX version ? which database ? which version of SqueakDBX/ConfigurationOf ?
do you have the dump or console output of the crash ?
thanks
Mariano
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
> > > On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote: > >> I get this error a lot when my system is under load. According to >> the pgsql docs, I should do this, "PQgetResult must be called >> repeatedly until it returns a null pointer, >> > > *indicating* that the *command* *is* done." >> >> > This is weird...it should not happen while using SqueakDBX. > If you see DBXConnection >> execute: > > just at the beginning we do: > > | aResult | > self isOpen ifFalse: > [ OpenDBXDriverError signal: 'You cannot execute a query if > the connection is not open' ]. > self releasePreviousResult. > > > that method #releasePreviousResult that exactly that. If you didn't > finish iterating all returns in your last query, we do it for you in the > next query. Of course, we also do that while closing the connection. > So...I cannot image how you can get that :( > Probably something with the pool and Glorp ? > > >> How do I handle this type of error in GlorpDBX? I'm using Pharo >> 1.2 and the Glorp/DBX build that was available when that was released. >> >> -- >> ~JT >> >> >> >> _______________________________________________ >> SqueakDBX mailing list >> SqueakDBX@lists.squeakfoundation.org >> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >> >> > > > -- > Mariano > http://marianopeck.wordpress.com > > > _______________________________________________ > SqueakDBX mailing list > SqueakDBX@lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx > >
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
THanks. Probably is not related...but did you read this post:
http://joachimtuchel.wordpress.com/2011/03/25/glorp-va-smalltalk-proxies-and...
cheers
mariano
On Wed, May 11, 2011 at 4:48 PM, John Toohey jt@parspro.com wrote:
- Pharo-1.1.1--Latest update: #11414
- Posstgres 8.3.7
- 32 bit compiled OpenDBX from approx 12 months ago
- Headless Linux server running on EC2
- VM is 15.2f
What I often get it the debugger firing with Glorp errors because the Glorp proxy does not instantiate an object associated with a foreign key. I can usually get out of those, but often in the morning, the VM itself has gone down. Attached is a PharoDebug from this morning.
On Wed, May 11, 2011 at 10:26, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Wed, May 11, 2011 at 3:54 PM, John Toohey jt@parspro.com wrote:
Well the VM crashes, but I can't be sure if its because of the resume or something else.
wow.... we will need more data in order to help there. Which Pharo/Squeak image, which VM ? which OS ? which OpenDBX version ? which database ? which version of SqueakDBX/ConfigurationOf ?
do you have the dump or console output of the crash ?
thanks
Mariano
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
> But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
> are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
> [ > ^aQuery executeWithParameters: #() in: self. > ] > on: GlorpDatabaseReadError > do: [:ex | Transcript show: ex messageText;cr. ex resume]. >
thanks
Mariano
> > On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < > marianopeck@gmail.com> wrote: > >> >> >> On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.comwrote: >> >>> I get this error a lot when my system is under load. According to >>> the pgsql docs, I should do this, "PQgetResult must be called >>> repeatedly until it returns a null pointer, >>> >> >> *indicating* that the *command* *is* done." >>> >>> >> This is weird...it should not happen while using SqueakDBX. >> If you see DBXConnection >> execute: >> >> just at the beginning we do: >> >> | aResult | >> self isOpen ifFalse: >> [ OpenDBXDriverError signal: 'You cannot execute a query if >> the connection is not open' ]. >> self releasePreviousResult. >> >> >> that method #releasePreviousResult that exactly that. If you didn't >> finish iterating all returns in your last query, we do it for you in the >> next query. Of course, we also do that while closing the connection. >> So...I cannot image how you can get that :( >> Probably something with the pool and Glorp ? >> >> >>> How do I handle this type of error in GlorpDBX? I'm using Pharo >>> 1.2 and the Glorp/DBX build that was available when that was released. >>> >>> -- >>> ~JT >>> >>> >>> >>> _______________________________________________ >>> SqueakDBX mailing list >>> SqueakDBX@lists.squeakfoundation.org >>> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >>> >>> >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com >> >> >> _______________________________________________ >> SqueakDBX mailing list >> SqueakDBX@lists.squeakfoundation.org >> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >> >> > > > -- > ~JT > > > > _______________________________________________ > SqueakDBX mailing list > SqueakDBX@lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx > >
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
Interesting, but it my cases, all the key does have values in the DB, yet Glorp sometimes returns a Proxy to my code, which then throws a DND exception when I access it as if its one of my objects.
On Wed, May 11, 2011 at 11:02, Mariano Martinez Peck marianopeck@gmail.comwrote:
THanks. Probably is not related...but did you read this post:
http://joachimtuchel.wordpress.com/2011/03/25/glorp-va-smalltalk-proxies-and...
cheers
mariano
On Wed, May 11, 2011 at 4:48 PM, John Toohey jt@parspro.com wrote:
- Pharo-1.1.1--Latest update: #11414
- Posstgres 8.3.7
- 32 bit compiled OpenDBX from approx 12 months ago
- Headless Linux server running on EC2
- VM is 15.2f
What I often get it the debugger firing with Glorp errors because the Glorp proxy does not instantiate an object associated with a foreign key. I can usually get out of those, but often in the morning, the VM itself has gone down. Attached is a PharoDebug from this morning.
On Wed, May 11, 2011 at 10:26, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Wed, May 11, 2011 at 3:54 PM, John Toohey jt@parspro.com wrote:
Well the VM crashes, but I can't be sure if its because of the resume or something else.
wow.... we will need more data in order to help there. Which Pharo/Squeak image, which VM ? which OS ? which OpenDBX version ? which database ? which version of SqueakDBX/ConfigurationOf ?
do you have the dump or console output of the crash ?
thanks
Mariano
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
> > > On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote: > >> But as these exceptions > > > which ones? I guess you are talking about GlorpDatabaseReadError > > >> are marked as Resumable, shouldn't it be enough for me to do :- > > > I am not sure if I understood correctly. Can you explain a little > more please? > > >> [ >> ^aQuery executeWithParameters: #() in: self. >> ] >> on: GlorpDatabaseReadError >> do: [:ex | Transcript show: ex messageText;cr. ex resume]. >> > > > > thanks > > Mariano > > > >> >> On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < >> marianopeck@gmail.com> wrote: >> >>> >>> >>> On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.comwrote: >>> >>>> I get this error a lot when my system is under load. According to >>>> the pgsql docs, I should do this, "PQgetResult must be called >>>> repeatedly until it returns a null pointer, >>>> >>> >>> *indicating* that the *command* *is* done." >>>> >>>> >>> This is weird...it should not happen while using SqueakDBX. >>> If you see DBXConnection >> execute: >>> >>> just at the beginning we do: >>> >>> | aResult | >>> self isOpen ifFalse: >>> [ OpenDBXDriverError signal: 'You cannot execute a query if >>> the connection is not open' ]. >>> self releasePreviousResult. >>> >>> >>> that method #releasePreviousResult that exactly that. If you >>> didn't finish iterating all returns in your last query, we do it for you in >>> the next query. Of course, we also do that while closing the connection. >>> So...I cannot image how you can get that :( >>> Probably something with the pool and Glorp ? >>> >>> >>>> How do I handle this type of error in GlorpDBX? I'm using Pharo >>>> 1.2 and the Glorp/DBX build that was available when that was released. >>>> >>>> -- >>>> ~JT >>>> >>>> >>>> >>>> _______________________________________________ >>>> SqueakDBX mailing list >>>> SqueakDBX@lists.squeakfoundation.org >>>> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >>>> >>>> >>> >>> >>> -- >>> Mariano >>> http://marianopeck.wordpress.com >>> >>> >>> _______________________________________________ >>> SqueakDBX mailing list >>> SqueakDBX@lists.squeakfoundation.org >>> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >>> >>> >> >> >> -- >> ~JT >> >> >> >> _______________________________________________ >> SqueakDBX mailing list >> SqueakDBX@lists.squeakfoundation.org >> http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx >> >> > > > -- > Mariano > http://marianopeck.wordpress.com > > > _______________________________________________ > SqueakDBX mailing list > SqueakDBX@lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx > >
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
in the image directory.. could be 3 logs file.
LowSpaceDebug.log PharoDebug.log ConflictChecker.log
those files. may be give to you some clue about the kind and source of problem
Best D.
On Wed, May 11, 2011 at 10:54 AM, John Toohey jt@parspro.com wrote:
Well the VM crashes, but I can't be sure if its because of the resume or something else.
On Tue, May 10, 2011 at 15:04, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 9:03 PM, John Toohey jt@parspro.com wrote:
The main one is the GlorpDatabaseReadError, and as its resumable, and a read error, I just want to retry it, until it succeeds. This is a production system, and having the debugger fire constantly, eventually messes up the entire image.
Ok....but if you resume... does it work finally ? I mean, if you do what you said, you fix your problem ?
thanks
On Tue, May 10, 2011 at 14:23, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 5:03 PM, John Toohey jt@parspro.com wrote:
But as these exceptions
which ones? I guess you are talking about GlorpDatabaseReadError
are marked as Resumable, shouldn't it be enough for me to do :-
I am not sure if I understood correctly. Can you explain a little more please?
[ ^aQuery executeWithParameters: #() in: self. ] on: GlorpDatabaseReadError do: [:ex | Transcript show: ex messageText;cr. ex resume].
thanks
Mariano
On Tue, May 10, 2011 at 10:39, Mariano Martinez Peck < marianopeck@gmail.com> wrote:
On Tue, May 10, 2011 at 4:31 PM, John Toohey jt@parspro.com wrote:
> I get this error a lot when my system is under load. According to the > pgsql docs, I should do this, "PQgetResult must be called repeatedly > until it returns a null pointer, >
*indicating* that the *command* *is* done." > > This is weird...it should not happen while using SqueakDBX. If you see DBXConnection >> execute:
just at the beginning we do:
| aResult | self isOpen ifFalse: [ OpenDBXDriverError signal: 'You cannot execute a query if
the connection is not open' ]. self releasePreviousResult.
that method #releasePreviousResult that exactly that. If you didn't finish iterating all returns in your last query, we do it for you in the next query. Of course, we also do that while closing the connection. So...I cannot image how you can get that :( Probably something with the pool and Glorp ?
> How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 > and the Glorp/DBX build that was available when that was released. > > -- > ~JT > > > > _______________________________________________ > SqueakDBX mailing list > SqueakDBX@lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx > >
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- Mariano http://marianopeck.wordpress.com
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
-- ~JT
SqueakDBX mailing list SqueakDBX@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/squeakdbx
Hi Norbert. John is having a problem with SqueakDBX and I just want to be sure it is what I think.... He gets the error "Another command is already in progress" and acoording to pgsql documentation "PQgetResult must be called repeatedly until it returns a null pointer, *indicating* that the *command* *is* done."
So...in the OpenDBX world, this would mean that he (SqueakDBX) didn't call
int odbx_result( odbx_t* handle, odbx_result_t** result, struct timeval* timeout, unsigned long chunk )
up to receive NULL ??
or this has nothing to do and it is another error??
thanks
mariano
---------- Forwarded message ---------- From: John Toohey jt@parspro.com Date: Tue, May 10, 2011 at 4:31 PM Subject: [SqueakDBX] OpenDBX: Another command is already in progress To: squeakdbx squeakdbx@lists.squeakfoundation.org
I get this error a lot when my system is under load. According to the pgsql docs, I should do this, "PQgetResult must be called repeatedly until it returns a null pointer, *indicating* that the *command* *is* done."
How do I handle this type of error in GlorpDBX? I'm using Pharo 1.2 and the Glorp/DBX build that was available when that was released.
squeakdbx@lists.squeakfoundation.org