| *** Mud has quit IRC | 01:40 | |
| *** gerhard has joined #ingres | 01:45 | |
| gerhard | Morning all | 01:46 |
|---|---|---|
| *** Mud has joined #ingres | 01:47 | |
| gerhard | Is there any CBF setting or trace point that will show the corresponding SQL in case of "escalating to table-level lock" situations? | 01:48 |
| *** Mud|afk2 has joined #ingres | 01:50 | |
| grantc | hi gerhard - there is | 01:51 |
| grantc | grep config.dat for "log_esc" | 01:52 |
| *** Mud has quit IRC | 01:53 | |
| gerhard | grantc, I think the log_esc params in config.dat just switch the lock escalation messages in errlog.log on or off... | 02:31 |
| gerhard | ...but I will not see the SQL that causes the lock escalation. | 02:32 |
| grantc | ah I see - you want the SQL that caused it not the tables upon which it's happenning | 02:32 |
| gerhard | yep | 02:32 |
| grantc | you might be able to extract the info from IMA ? | 02:33 |
| *** KermitTheFragger has joined #ingres | 02:33 | |
| gerhard | you mean something like "select * from ima_server_sessions..." ? | 02:34 |
| gerhard | BTW, my customers machine (Win-32, 9.2.0, P13354) has all log_esc params turned OFF in the config.dat and nevertheless, lock escalation messages are shown in the errlog.log | 02:36 |
| gerhard | Is that a bug? | 02:37 |
| grantc | if the change was made post server start up then no - the errlog.log will tell you what the server started up with. If Ingres has been restarted then I guess it is a bug | 02:43 |
| gerhard | Then it is a bug because I have definitely restarted Ingres after last cbf change. | 03:11 |
| *** withdefault has joined #ingres | 04:15 | |
| *** ChanServ sets mode: +o withdefault | 04:15 | |
| *** withdefault sets mode: -o withdefault | 04:15 | |
| *** rossand has joined #ingres | 04:19 | |
| *** ChanServ sets mode: +o rossand | 04:19 | |
| *** zxiiro has quit IRC | 04:34 | |
| *** zxiiro has joined #ingres | 05:14 | |
| *** ChanServ sets mode: +v zxiiro | 05:14 | |
| *** Mud|afk2 has quit IRC | 05:42 | |
| *** Mud has joined #ingres | 05:46 | |
| Vroomfondle | hrm | 06:04 |
| Vroomfondle | grantc: Error: 7 - Unable to close non-result resource statement (when executing ingres_query) | 06:04 |
| Vroomfondle | any idea what the cause of that might be? Code worked before we moved over to the new version of the driver. I think it's actually a warning rather than an error, but I still need to fix it somehow | 06:05 |
| *** mull has joined #ingres | 06:05 | |
| Vroomfondle | I think it's related to executing prepared statements, as the majority of our code (which uses an older mechanism) is working fine | 06:08 |
| grantc | it's doing clean up from an earlier statement - you may or may not be able to ignore it. Can you generate an API trace with the problem code? | 06:20 |
| Vroomfondle | I could if I knew how | 06:23 |
| awatkins | grantc: how do you generate an api trace from within the ingres code? | 06:23 |
| Vroomfondle | thing is, I can't ignore it because my own custom application-wide error handler doesn't let me ;) | 06:24 |
| Vroomfondle | <-- bit of a nazi regarding errors/warnings ;) | 06:25 |
| grantc | http://community.ingres.com/w/files/3/3b/GettingStartedwithIngresandPHP.pdf - Slide 40 | 06:26 |
| grantc | awatkins: it's done using environment variables that have to be set before hand | 06:26 |
| * Vroomfondle takes a look | 06:28 | |
| grantc | http://docs.google.com/present/edit?id=0Ac-MQvdi6eYBZGdmYnMyOGpfNDVka3Fxc25ncQ&hl=en as well | 06:31 |
| awatkins | cool. I was getting no common charset errors, and had to go mucking around the codebase to find which env variable to set to get it to work. wondered how to activate the tracing so it'd make my life easier | 06:33 |
| grantc | lunchtime ... | 06:36 |
| *** grantc is now known as grantc_fooooood | 06:36 | |
| Vroomfondle | grantc_fooooood: k, cheers, will ask my boss to add the necessary lines to our config and then let you know the result | 06:38 |
| *** cthibert has joined #ingres | 06:41 | |
| *** Mud has quit IRC | 06:54 | |
| DarylM | Guten morgen all, awatkins. How goes the ruby fun? | 07:00 |
| Vroomfondle | hrm | 07:04 |
| Vroomfondle | grantc_fooooood: We've added the requisite environment vars to our apache config (as SetEnv statements) and restarted Apache. It doesn't seem to be logging a trace anywhere though... | 07:05 |
| Vroomfondle | err, wait | 07:06 |
| * Vroomfondle greps the conf file which his boss is supposed to have edited | 07:06 | |
| * Vroomfondle wanders off to his boss with a big question-mark over his head | 07:06 | |
| Vroomfondle | ah. I asked him to apply it to our dev server. He actually applied it to our *live* server. | 07:08 |
| Vroomfondle | That's not the kind of mistake one really wants to make! ;) | 07:09 |
| *** grantc_fooooood is now known as grantc | 07:19 | |
| grantc | err no not really since it will hit performance a bit | 07:19 |
| Vroomfondle | now applied to dev, but... erm, still doesn't seem to be working | 07:22 |
| *** gerhard has quit IRC | 07:23 | |
| *** Mud has joined #ingres | 07:23 | |
| Vroomfondle | SetEnv II_API_TRACE 5 | 07:23 |
| Vroomfondle | SetEnv II_API_LOG /www/logs/ingres_api_trace.log | 07:23 |
| Vroomfondle | ...right? | 07:23 |
| grantc | you might need to set them in the script that starts Apache and use PassEnv to make sure the PHP engine can see the variables | 07:28 |
| *** awatkins has left #ingres | 07:32 | |
| Vroomfondle | seems unlikely as we use SetEnvs (in the same part of the config file) to set II_SYSTEM... | 08:02 |
| grantc | are you using mod_php? | 08:09 |
| *** rossand has quit IRC | 08:37 | |
| *** rossand has joined #ingres | 08:38 | |
| *** ChanServ sets mode: +o rossand | 08:39 | |
| Vroomfondle | grantc: yes | 08:50 |
| grantc | uh ok - i've never got SetEnv to work, perhaps it's time to try again... | 08:50 |
| Vroomfondle | heh | 08:51 |
| Vroomfondle | it's entirely possible that ours don't actually work and we haven't noticed cos II_SYSTEM is getting set somewhere else | 08:51 |
| Vroomfondle | that sort of thing happens a lot around here. | 08:51 |
| grantc | :D | 08:51 |
| pboro | :D | 08:53 |
| grantc | I've just tried it again and nope - it does not work for me... | 09:00 |
| Vroomfondle | I'll dig around a bit more tomorrow | 09:00 |
| grantc | if you can send me your httpd.conf that would be useful (grantc@php.net or my ingres.com address) | 09:02 |
| pboro | you can easily check whether SetEnv works with PHP by creating a php script containing <?php phpinfo(); ?> | 09:02 |
| pboro | then simply execute it thru' web and check the env variables list :) | 09:04 |
| *** withdefault has quit IRC | 09:05 | |
| grantc | it's there in the $_server[] list not in the environment variable list | 09:05 |
| grantc | at least that's what i see | 09:05 |
| pboro | then it prolly doesn't work as expected... | 09:06 |
| pboro | I guess it only affects cgi's etc, lemme check the manual | 09:07 |
| pboro | "Sets an environment variable, which is then passed on to CGI scripts and SSI pages." | 09:08 |
| pboro | doesn't look too promising | 09:08 |
| grantc | err no | 09:09 |
| grantc | So if you call PHP via the CGI module then you are ok but mod_php is a different story | 09:10 |
| pboro | I guess it could be useful that the ingres driver module also considered the values in the $_server[] list... but in the end, you still need to define the LD_LIBRARY_PATH elsewhere | 09:11 |
| grantc | I've thought about that but it would only be specific to apache not to other servers such as IIS/lighttpd | 09:13 |
| pboro | hmm IIS and lighttpd don't use $_server at all? | 09:14 |
| pboro | but setting such values might be impossible.. yeah | 09:15 |
| grantc | not sure, i guess so but there are apache specific functions for fetching setenv'd variables | 09:15 |
| pboro | yup | 09:15 |
| pboro | yea but I didn't mean implementing specific stuff for setenv but simply looking up $_server[] | 09:16 |
| pboro | but there's no much point anyway so :) | 09:16 |
| * grantc picks up his extending and embedding php book | 09:17 | |
| pboro | :D | 09:17 |
| pboro | damn I hate when my laptop starts failing wlan connections | 09:17 |
| grantc | someone playing with a microwave? | 09:18 |
| pboro | naah, something wrong with the wlan mini-pci card | 09:18 |
| grantc | http://xkcd.com/654/ | 09:19 |
| pboro | haha | 09:19 |
| *** grantc has quit IRC | 09:46 | |
| *** KermitTheFragger has quit IRC | 09:53 | |
| *** Mud has quit IRC | 10:14 | |
| *** Mud has joined #ingres | 11:04 | |
| *** Alex| has joined #ingres | 12:02 | |
| *** ChanServ sets mode: +o Alex| | 12:02 | |
| *** Mud has quit IRC | 12:09 | |
| *** Mud has joined #ingres | 12:15 | |
| *** zxiiro has quit IRC | 12:20 | |
| *** Eugene_K has joined #ingres | 12:41 | |
| Eugene_K | Can someone remind me - which 3 Windows security rights I need to assign to the Windows user (which will install Ingres)? | 12:44 |
| Eugene_K | Oh, yes, Hello All | 12:44 |
| Eugene_K | I have installed clean Ingres 9.2.1 and my rmcmd server will not start. Errors (in errlog.log): E_GC0153_GCN_SRV_STARTUPServer Registration: class RMCMD, address II\RMCMD\1894 | 12:50 |
| Eugene_K | Any suggestions? | 12:50 |
| DarylM | I always have to hit the Windows specific documentation for that. "Act as admin" I think is one of them | 12:56 |
| Eugene_K | Yeh, Daryl, spent about 12 minutes looking for that info - nowhere to be found... | 12:56 |
| Eugene_K | One is definitely - Act as part of Operating System | 12:57 |
| DarylM | I only have the Linux manuals easily at hand... :-) | 12:58 |
| Eugene_K | Thanks for trying Daryl | 12:59 |
| Eugene_K | I know you are Linux person : -) | 13:00 |
| *** Eugene_K has quit IRC | 13:05 | |
| *** zxiiro has joined #ingres | 13:07 | |
| *** ChanServ sets mode: +v zxiiro | 13:07 | |
| *** Mud has quit IRC | 13:59 | |
| *** cthibert has left #ingres | 14:03 | |
| *** Alex| has quit IRC | 14:06 | |
| *** HY_ has joined #ingres | 14:37 | |
| *** mull has quit IRC | 15:59 | |
| *** rossand has quit IRC | 18:16 | |
| *** cytrinox_ has joined #ingres | 19:20 | |
| *** cytrinox has quit IRC | 19:24 | |
| *** cytrinox_ is now known as cytrinox | 19:24 | |
| *** mull has joined #ingres | 19:36 | |
| *** mull has quit IRC | 19:58 | |
| *** Alex| has joined #ingres | 22:38 | |
| *** ChanServ sets mode: +o Alex| | 22:38 | |
| *** Alex| has quit IRC | 23:04 | |
| *** Alex| has joined #ingres | 23:25 | |
| *** ChanServ sets mode: +o Alex| | 23:25 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!