I’m testing 3.4.2rc2 on FreeBSD 5-Stable with PostGreSQL 8.02, and I
have lot of problems using DBD:Pg-1.41
The most visible symptom is that URIs like:
http://rt.mydomain.it/Ticket/Display.html?id=2
work less than once every three clicks (returning 'No ticket specified’
on the failed attempts)
Reverting DBD:Pg to version 1.32 fixed these problems.
I know Jesse reported this as a DBD:Pg fault, but is there nothing that
could be tried on the code side to work around these incompatibilities?
Any other reports ?
Angelo Turetta
I’m testing 3.4.2rc2 on FreeBSD 5-Stable with PostGreSQL 8.02, and I
have lot of problems using DBD:Pg-1.41
The most visible symptom is that URIs like:
http://rt.mydomain.it/Ticket/Display.html?id=2
work less than once every three clicks (returning ‘No ticket specified’
on the failed attempts)
How about logs? Any more detail about what’s going wrong?
I know Jesse reported this as a DBD:Pg fault, but is there nothing that
could be tried on the code side to work around these incompatibilities?
The DBD::Pg fault was an only an issue issue on Pg 7.x, not 8.x. But
no, there was nothing that we could do. It broke all sorts of
applications.
Jesse Vincent wrote:
I’m testing 3.4.2rc2 on FreeBSD 5-Stable with PostGreSQL 8.02, and I
have lot of problems using DBD:Pg-1.41
The most visible symptom is that URIs like:
http://rt.mydomain.it/Ticket/Display.html?id=2
work less than once every three clicks (returning ‘No ticket specified’
on the failed attempts)
How about logs? Any more detail about what’s going wrong?
I know Jesse reported this as a DBD:Pg fault, but is there nothing that
could be tried on the code side to work around these incompatibilities?
The DBD::Pg fault was an only an issue issue on Pg 7.x, not 8.x. But
no, there was nothing that we could do. It broke all sorts of
applications.
I think it may be a bug in 3.4.2.
I’m using exactly the same combination (FreeBSD, Postgres 8.0.2) - but
with RT3.4.1.
I tried 3.4.2RC1, but I got the same problems as Angelo and reverted
back to 3.4.1
Can you install 3.4.1 and see if your problems go away?
I didn’t see anything in the logs that would have indicated an error.
cheers,
Rainer
~ Rainer Duffner - rainer@ultra-secure.de ~
~ Freising - Munich - Germany ~
~ Unix - Linux - BSD - OpenSource - Security ~
~ http://www.ultra-secure.de/~rainer/pubkey.pgp ~
I think it may be a bug in 3.4.2.
I’m using exactly the same combination (FreeBSD, Postgres 8.0.2) - but
with RT3.4.1.
I tried 3.4.2RC1, but I got the same problems as Angelo and reverted
back to 3.4.1
I’ve tried to replicate this on RT 3.4.2 + DBD::Pg 1.41 + Pg 8.02. And I
can’t. I’m afraid I can’t do much more without more debugging
information from y’all.
Jesse Vincent wrote:
I think it may be a bug in 3.4.2.
I’m using exactly the same combination (FreeBSD, Postgres 8.0.2) - but
with RT3.4.1.
I tried 3.4.2RC1, but I got the same problems as Angelo and reverted
back to 3.4.1
I’ve tried to replicate this on RT 3.4.2 + DBD::Pg 1.41 + Pg 8.02. And I
can’t. I’m afraid I can’t do much more without more debugging
information from y’all.
That’s what I was suspecting: it’s a misconfiguration (of my origin)
somehow.
I’ve customized my RT a bit, but only with mods from the wiki and mail-list.
I’ll see what I can do.
cheers,
Rainer
~ Rainer Duffner - rainer@ultra-secure.de ~
~ Freising - Munich - Germany ~
~ Unix - Linux - BSD - OpenSource - Security ~
~ http://www.ultra-secure.de/~rainer/pubkey.pgp ~
Jesse Vincent wrote:> On Sat, May 07, 2005 at 06:54:24PM +0200, Rainer Duffner wrote:
I think it may be a bug in 3.4.2.
I’m using exactly the same combination (FreeBSD, Postgres 8.0.2) - but
with RT3.4.1.
I tried 3.4.2RC1, but I got the same problems as Angelo and reverted
back to 3.4.1
I’ve tried to replicate this on RT 3.4.2 + DBD::Pg 1.41 + Pg 8.02. And I
can’t. I’m afraid I can’t do much more without more debugging
information from y’all.
I had some more time to experiment, and I can confirm that 3.4.2 has
some serious regression on FreeBSD/Apache2/mod_perl2(2.0rc3)/PostgreSql
compared to 3.4.1, and DBD:Pg doesn’t change much the result (in fact,
with 1.32 some of the symptoms are somewhat attenuated, but the system
is still unusable)
All dependencies are installed with the FreeBSD ports system, on
5.3-STABLE (please find attached the version of all deps)
In my setup, 3.4.1 works just fine. If I install 3.4.2 (changing nothing
else), all kinds of problems show up: I list some of them from memory
-
In the query builder the ‘Add’ button in the ‘Add Criteria’ section
works only the first time after restarting apache, and then does nothing
-
In the query builder the ‘Search’ button does nothing most of the
times. Occasionally it starts the search :-).
-
In the query builder ‘New Query’ does nothing
-
clicking on a ticket’s link (search results or Home page), you get the
ticket data only if you leave the browser quiet for some time. (if you
click on two tickets one after the other, the second displays ‘No ticket
specified’)
-
after you log out, you cannot log in any more (until apache is restarted)
-
the overall impression is as if pages were being ‘cached’.
I tried both Firefox and Internet Explorer (on which I set the cache to
verify the page every visit) with the same results.
The log file (‘debug’) is not terribly useful, but I attach it anyway.
Please tell me how can I help some more with the debug (I may even
provide a shell access to my test system if you cannot reproduce the error).
To setup a test machine, you may wish to download the port I’ve created
(I’m using it to install RT) from
http://mail.bestunion.it/rt342-port.tgz
please untar in /usr/ports (port tree should be newer than 2005-03-10), then
cd www/rt34 && make -DWITH_APACHE2 DB_TYPE=Pg install
Thanks
Angelo
deps.txt (2.83 KB)
rt20050513.log.bz2 (1015 Bytes)
[I’ve removed rt-users from the CC list]
I had some more time to experiment, and I can confirm that 3.4.2 has
some serious regression on FreeBSD/Apache2/mod_perl2(2.0rc3)/PostgreSql
compared to 3.4.1, and DBD:Pg doesn’t change much the result (in fact,
with 1.32 some of the symptoms are somewhat attenuated, but the system
is still unusable)
Can you try with mod_perl 1, FastCGI or standalone_httpd? I’d like to
try to nail down which component is causing the issue. mod_perl2 is so
incredibly not ready for prime-time. (Especially given the major
incompatible changes in the last release candidate)
Jesse
Jesse Vincent wrote:
[I’ve removed rt-users from the CC list]
I had some more time to experiment, and I can confirm that 3.4.2 has
some serious regression on FreeBSD/Apache2/mod_perl2(2.0rc3)/PostgreSql
compared to 3.4.1, and DBD:Pg doesn’t change much the result (in fact,
with 1.32 some of the symptoms are somewhat attenuated, but the system
is still unusable)
Can you try with mod_perl 1, FastCGI or standalone_httpd? I’d like to
try to nail down which component is causing the issue. mod_perl2 is so
incredibly not ready for prime-time. (Especially given the major
incompatible changes in the last release candidate)
Jesse
FastGCI with Apache2 works perfectly. I may test standalone as an
academic exercise, but I won’t promise…
Now it’s just a matter of making it work with the newest rc >3 
Seriously, I think it’s just a waste to try and debug the current code
on mod_perl2rc3. Much better to spend the time adapting RT to the new
API in rc4 and subsequent. Just put somewhere a disclaimer saying that
3.4.2 is not working with mod_perl2.
Angelo.
FastGCI with Apache2 works perfectly. I may test standalone as an
academic exercise, but I won’t promise…
Now it’s just a matter of making it work with the newest rc >3
Seriously, I think it’s just a waste to try and debug the current code
on mod_perl2rc3. Much better to spend the time adapting RT to the new
API in rc4 and subsequent. Just put somewhere a disclaimer saying that
3.4.2 is not working with mod_perl2.
There’s a reason I’ve been unwilling to support mod_perl2 before its
release. 
How’s this language?
mod_perl 2.0 isn't quite ready for prime_time just yet;
We _strongly_ recommend that you use Apache 1.3 or
FastCGI.
Jesse
Jesse Vincent wrote:
FastGCI with Apache2 works perfectly.
Good.
Thanks a lot.
I may test standalone as an
academic exercise, but I won’t promise…
Now it’s just a matter of making it work with the newest rc >3
Seriously, I think it’s just a waste to try and debug the current code
on mod_perl2rc3. Much better to spend the time adapting RT to the new
API in rc4 and subsequent. Just put somewhere a disclaimer saying that
3.4.2 is not working with mod_perl2.
OK, using mod_perl2 probably wasn’t that smart in the first place.
There’s a reason I’ve been unwilling to support mod_perl2 before its
release.
How’s this language?
mod_perl 2.0 isn't quite ready for prime_time just yet;
We _strongly_ recommend that you use Apache 1.3 or
FastCGI.
I modified ManualInstallation in the wiki a bit to that tune.
Rainer
~ Rainer Duffner - rainer@ultra-secure.de ~
~ Unix - Linux - BSD - OpenSource - Security ~
~ http://www.ultra-secure.de/~rainer/pubkey.pgp ~
How’s this language?
mod_perl 2.0 isn't quite ready for prime_time just yet;
We _strongly_ recommend that you use Apache 1.3 or
FastCGI.
I modified ManualInstallation in the wiki a bit to that tune.
Thanks. That’s a verbatim quote from the existing README