Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-29 Thread Jean Wainer
On Tue, 29 Jun 2004 08:34:30 +0700
François Wautier [EMAIL PROTECTED] wrote:

 Hi,
 
 The patch applies does not solve the underlying problem, the fact that you 
 cannot connect to the DB server. It just makes sure the system behaves  properly 
 even when one cannot access the DB.

I see. You said, in your previous message:
The code in vpopmail was first trying to connect to the server for a given database. 
If that was failing, it would try the server only and if that succeeded, it was 
considering that the server was fine but that the DB was not created.

So, wouldnt it be safer to consider the second try, if it succeeds? Sorry, I dont 
understand why consider that the database was not created only because it took 2 tries 
to connect to the database.

 In the case of email delivery, proper behaviour means that the message is put 
 back into the queue for later delivery.
Right.

 In other cases, proper behaviour may means that the system complains it does 
 not know your user. I do not know.
Yes, that's actually whats happening. Specially for Outlook users who keep getting a 
request for their password (using Sylpheed, in my case, I can see the database does 
not exist error).

 Now that I think about it, that could be the problem I am having. When using 
 IMP-Courier imap-vpopmail  some of my users need to login many times. It 
 could be that, because of the problem, authentication fails and they have to 
 loggin again.
Yes, it happes with me using courier + squirrelmail too.


thanks!

--Jw.
 
 Best Regards,
   François
 
 
 
 
 l
  On Mon, 28 Jun 2004 21:54:55 +0700
 
  François Wautier [EMAIL PROTECTED] wrote:
   Hi,
  
   Since I submitted that patch.. I guess I should answer.
  
   The code in vpopmail was first trying to connect to the server for a
   given database. If that was failing, it would try the server only and if
   that succeeded, it was considering that the server was fine but that the
   DB was not created.
  
   Apparently, under some load, it could happen that the first request would
   fail, but the second would be accepted...
  
   All the patch does is make the code connect to the server first and then
   use the Database. If the first call fails, the delivery fails (temporary
   failure) and the message is delivered later.
  
   Cheers,
 François
 
  Thanks for the explanation, François :)
 
  I was having this problem, not only when delivering messages, but specially
  when checking POP3 mail (vchkpw). Will that be fixed too? Or am I going to
  get a different error? Also, using Antonio Nati's qmail-smtpd-chkusr patch
  (as in http://www.interazioni.it/qmail/qmail-smtpd-chkusr-10.html), what
  will be the effect with the new corrected mysql patch? I was getting some
  no such user errors before...
 
  thanks!
 
  Jean
 


Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-29 Thread François Wautier
Hi,

 I see. You said, in your previous message:
 The code in vpopmail was first trying to connect to the server for a given
 database. If that was failing, it would try the server only and if that
 succeeded, it was considering that the server was fine but that the DB was
 not created.

 So, wouldnt it be safer to consider the second try, if it succeeds? Sorry,
 I dont understand why consider that the database was not created only
 because it took 2 tries to connect to the database.

The way the program before patch was working was

if(openDB( Host, Database) != OK) {
if(openDB(Host) == OK) {
create database
}
}

After the patch it is

if(openDB(Host) != OK) {
return Can not connect
}

if(useDB(Database) != OK) {
create database
}

A trivial patch really.

In the pre-patch case, if the first openDB fails, it could be because the DB 
is not available...And it goes downhill from there

Best Regards,
François


Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-28 Thread Jean Wainer
On Fri, 25 Jun 2004 18:17:53 -0700
Tom Collins [EMAIL PROTECTED] wrote:

 http://vpopmail.sf.net
 
 Release Notes:
 
 There are significant changes in here for MySQL and Postgres backends.
 
 If you've had errors stating couldn't create table/database because it
 already exists with MySQL, you should definitely upgrade.

Well, I was having such errors with 5.4.3, and I did upgrade. It seems to be working 
fine, now.
I noticed that this error occured only with the server under heavy load (over 8, 10). 
Since it is fixed now in 5.4.5, I have a question out of curiosity: What was the bug 
that was causing this error?

Thanks,

Jean


Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-28 Thread François Wautier
Hi,

Since I submitted that patch.. I guess I should answer.

The code in vpopmail was first trying to connect to the server for a given 
database. If that was failing, it would try the server only and if that 
succeeded, it was considering that the server was fine but that the DB was 
not created.

Apparently, under some load, it could happen that the first request would 
fail, but the second would be accepted... 

All the patch does is make the code connect to the server first and then use 
the Database. If the first call fails, the delivery fails (temporary failure) 
and the message is delivered later.

Cheers,
François


 On Fri, 25 Jun 2004 18:17:53 -0700

 Tom Collins [EMAIL PROTECTED] wrote:
  http://vpopmail.sf.net
 
  Release Notes:
 
  There are significant changes in here for MySQL and Postgres backends.
 
  If you've had errors stating couldn't create table/database because it
  already exists with MySQL, you should definitely upgrade.

 Well, I was having such errors with 5.4.3, and I did upgrade. It seems to
 be working fine, now. I noticed that this error occured only with the
 server under heavy load (over 8, 10). Since it is fixed now in 5.4.5, I
 have a question out of curiosity: What was the bug that was causing this
 error?

 Thanks,

 Jean


Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-28 Thread Jean Wainer
On Mon, 28 Jun 2004 21:54:55 +0700
François Wautier [EMAIL PROTECTED] wrote:

 Hi,
 
 Since I submitted that patch.. I guess I should answer.
 
 The code in vpopmail was first trying to connect to the server for a given 
 database. If that was failing, it would try the server only and if that 
 succeeded, it was considering that the server was fine but that the DB was 
 not created.
 
 Apparently, under some load, it could happen that the first request would 
 fail, but the second would be accepted... 
 
 All the patch does is make the code connect to the server first and then use 
 the Database. If the first call fails, the delivery fails (temporary failure) 
 and the message is delivered later.
 
 Cheers,
   François


Thanks for the explanation, François :)

I was having this problem, not only when delivering messages, but specially when 
checking POP3 mail (vchkpw). Will that be fixed too? Or am I going to get a different 
error?
Also, using Antonio Nati's qmail-smtpd-chkusr patch (as in 
http://www.interazioni.it/qmail/qmail-smtpd-chkusr-10.html), what will be the effect 
with the new corrected mysql patch? I was getting some no such user errors before...

thanks!

Jean


Re: [vchkpw] Vpopmail 5.4.5 released

2004-06-28 Thread François Wautier
Hi,

The patch applies does not solve the underlying problem, the fact that you 
cannot connect to the DB server. It just makes sure the system behaves 
properly even when one cannot access the DB.

In the case of email delivery, proper behaviour means that the message is put 
back into the queue for later delivery.  

In other cases, proper behaviour may means that the system complains it does 
not know your user. I do not know.

Now that I think about it, that could be the problem I am having. When using 
IMP-Courier imap-vpopmail  some of my users need to login many times. It 
could be that, because of the problem, authentication fails and they have to 
loggin again.

Best Regards,
François




l
 On Mon, 28 Jun 2004 21:54:55 +0700

 François Wautier [EMAIL PROTECTED] wrote:
  Hi,
 
  Since I submitted that patch.. I guess I should answer.
 
  The code in vpopmail was first trying to connect to the server for a
  given database. If that was failing, it would try the server only and if
  that succeeded, it was considering that the server was fine but that the
  DB was not created.
 
  Apparently, under some load, it could happen that the first request would
  fail, but the second would be accepted...
 
  All the patch does is make the code connect to the server first and then
  use the Database. If the first call fails, the delivery fails (temporary
  failure) and the message is delivered later.
 
  Cheers,
  François

 Thanks for the explanation, François :)

 I was having this problem, not only when delivering messages, but specially
 when checking POP3 mail (vchkpw). Will that be fixed too? Or am I going to
 get a different error? Also, using Antonio Nati's qmail-smtpd-chkusr patch
 (as in http://www.interazioni.it/qmail/qmail-smtpd-chkusr-10.html), what
 will be the effect with the new corrected mysql patch? I was getting some
 no such user errors before...

 thanks!

 Jean


[vchkpw] Vpopmail 5.4.5 released

2004-06-25 Thread Tom Collins
http://vpopmail.sf.net
Release Notes:
There are significant changes in here for MySQL and Postgres backends.
If you had problems with Postgres and roaming users, you should
definitely upgrade.
If you've had errors stating couldn't create table/database because it
already exists with MySQL, you should definitely upgrade.
ChangeLog:
fernando (at) telemacro (dot) com (dot) br
- Patch for vpgsql.c fixes bug with Postgres and roaming users
  (POP before SMTP). [895501]
Françoi Wautier
- Fix method used to open database in vauth_open_update of
  vmysql.c. [967994, 946983]
Pit Palme
- Show 'delete' as valid option to vdelivermail in docs. [951245]
rstml
- Hide error message during POP3 auth with Postgres. [915485]
Tom Collins
- Fix `vuserinfo -l` output, based on Bill Shupp's patch
  (moved code to a single function call). [961742]