*** toumi01 has quit IRC | 00:09 | |
*** PaulM05 has joined #ingres | 00:36 | |
*** grantc has quit IRC | 00:43 | |
*** withdefault has joined #ingres | 00:53 | |
*** ChanServ sets mode: +o withdefault | 00:53 | |
*** grantc has joined #ingres | 00:57 | |
*** KermitTheFragger has joined #ingres | 01:04 | |
*** grantc has quit IRC | 01:05 | |
*** grantc has joined #ingres | 01:19 | |
Dejan | morning guys | 01:29 |
---|---|---|
PaulM05 | hi | 01:30 |
Vroomfondle | hello. | 01:34 |
*** grantc has quit IRC | 01:34 | |
*** grantc has joined #ingres | 01:34 | |
*** Mud has joined #ingres | 02:02 | |
pboro | where does archiver store its position regarding to logs? | 02:20 |
pboro | I'm facing E_DM9838_ARCH_DB_BAD_LA | 02:20 |
PaulM05 | there's a last log address journalled in the database cnf file - see infodb output | 02:23 |
pboro | how do I know which database? :D | 02:31 |
pboro | maybe iiacp.log... lemme check | 02:31 |
pboro | not good, it's iidbdb | 02:31 |
pboro | Last Log Address Journaled : <1230534651:60775:208> | 02:32 |
pboro | ah... journal is not in sync because it was copied over from yesterday | 02:33 |
pboro | is there a way to reset it? | 02:33 |
PaulM05 | copied from where? | 02:33 |
pboro | from the same installation but a copy from few hours before | 02:34 |
pboro | admin did live upgrade on solaris and apparently it messed up something with the journals | 02:34 |
PaulM05 | well first thing I would do is check whether the database itself is ok | 02:35 |
PaulM05 | is the system in use? | 02:35 |
pboro | it is | 02:35 |
pboro | no errors are coming out and I can query catalogs in iidbdb without errors | 02:35 |
pboro | I'm pretty certain that the db is ok | 02:35 |
PaulM05 | in which case the best thing to do is probably disable journaling on iidbdb and then re-enable with a new checkpoint | 02:36 |
PaulM05 | you'll need a 5min window of exclusive access to iidbdb | 02:36 |
pboro | ok, disable it using... alterdb -disable_journaling? or how | 02:36 |
PaulM05 | yes alterdb | 02:37 |
pboro | why 5min window of excl access? :) if I just do alterdb -disable_journaling and then run ckpdb? or what was the idea? | 02:37 |
PaulM05 | you'll need to do ckpdb +j to re-enable journalling and that will take a database lock | 02:38 |
pboro | ok, that's not a prob and the checkpoint will prolly be a very quick one | 02:38 |
PaulM05 | that should be ok but you won't get any new connections to any other database | 02:38 |
PaulM05 | yes iidbdb is typically small which is why I said 5min | 02:38 |
pboro | I guess I need to start archiver after running alterdb? ie run "ingstart -dmfacp"? | 02:39 |
pboro | and then ckpdb +j? | 02:39 |
PaulM05 | yes | 02:39 |
pboro | ok, lemme try :D | 02:39 |
pboro | ok, worked | 02:40 |
pboro | thanks! | 02:40 |
PaulM05 | no problem | 02:40 |
*** atrofast has joined #ingres | 03:09 | |
grantc | Vroomfondle, i've managed to reproduce your problem | 03:19 |
pboro | apparently if a database columns contains \r\n on Solaris, report writer and qbf fails to output the text correctly | 03:35 |
pboro | it seems that the string just gets cut at the first \r | 03:36 |
pboro | is this some known problem, any tips? :) | 03:36 |
pboro | apparently it's not possible to get rows containing \n or \r with WHERE col LIKE | 03:39 |
pboro | at least WHERE col LIKE '\n' or LIKE X'0A' does not appear to work as expected | 03:39 |
Vroomfondle | grantc: woohoo! | 03:40 |
Vroomfondle | that's what I like to hear. | 03:40 |
grantc | i can only reproduce it on linux | 03:40 |
grantc | solaris even ... | 03:40 |
pboro | oops, my mistake :) | 03:41 |
pboro | LIKE works as should | 03:41 |
grantc | line 6925 "value_long = (long) *((II_INT4 *) columnData->dv_value);" is failing | 03:41 |
pboro | but reports and qbf fails on solaris with \r's | 03:41 |
pboro | we'll remove \r's but it's still weird | 03:41 |
PaulM05 | not come across that one pboro | 03:43 |
pboro | I'll have to put up a test case and try it on different servers | 03:44 |
*** Mud has quit IRC | 03:53 | |
*** Mud has joined #ingres | 03:57 | |
pboro | naah, report writer works fine, prolly something weird in the report then | 04:03 |
*** atrofast has quit IRC | 04:13 | |
*** atrofast has joined #ingres | 04:14 | |
*** cthibert has joined #ingres | 04:14 | |
pboro | ahh, figures... report writer only outputs first four rows :) | 04:26 |
pboro | I wonder if it's possible to make it dynamic, ie print as long as there's something to print | 04:27 |
pboro | ie if I have col with value 'first\nsecond\nthird\nfourth\nfifth\nsixth", report writer only outputs first four rows when using .PRINT (c256) | 04:29 |
*** zxiiro has quit IRC | 05:02 | |
*** gerhard has joined #ingres | 05:05 | |
*** PaulM05 is now known as PaulM05_lunch | 05:10 | |
*** zxiiro has joined #ingres | 05:45 | |
*** ChanServ sets mode: +v zxiiro | 05:45 | |
*** rossand has joined #ingres | 05:48 | |
*** ChanServ sets mode: +o rossand | 05:48 | |
*** PaulM05_lunch is now known as PaulM05 | 06:18 | |
gerhard | Hi all | 06:25 |
PaulM05 | hi gerhard | 06:26 |
atrofast | Does anyone know what the AS BEGIN EXECUTE INTERNAL; means for the procedures defined in src/dbutil/duf/duc/ducdata.c? | 06:30 |
withdefault | @atrofast, found this in the parser code | 06:32 |
ii_log | withdefault: Error: "atrofast," is not a valid command. | 06:32 |
PaulM05 | it means that there's a hook into the DBMS server and the procedure exists as a way of calling it from embedded SQL - it's used for things like verifydb, destroydb and createdb | 06:32 |
atrofast | Ah so where are those procedures actually defined then? :) | 06:33 |
PaulM05 | I'm just trying to remind myself of that | 06:34 |
atrofast | I'm looking to add a couple of new procedures and am uncertain where to put them if not in ducdata.c | 06:35 |
PaulM05 | what sort of procedures? | 06:35 |
withdefault | atrofast, comments acording to parser http://lxr.ingres.com/lxr/source/src/back/psf/psl/pslsgram.yi#13200 | 06:35 |
PaulM05 | have a look at qeaiproc.c in qeq | 06:37 |
atrofast | Thanks withdefault! | 06:37 |
atrofast | PaulM05: that adds and removes rows from a new system catalog | 06:37 |
*** toumi01 has joined #ingres | 06:38 | |
PaulM05 | you shouldn't need to do an execute internal for that I don't think | 06:38 |
atrofast | Oh that's not going to work | 06:38 |
atrofast | Since I'd like other users to be able to use it than $ingres | 06:39 |
atrofast | Is it even possible to create a stored proc that any user can run that can edit a system catalog? | 06:40 |
PaulM05 | I think if you create it as a regular procedure - using SQL inserts/updates etc - then the user running it needs permissions on the tables concerned - in this case the system catalog | 06:44 |
atrofast | Can I make an iiblah_catalog public? | 06:44 |
atrofast | Or is Ingres going to disallow that? | 06:44 |
PaulM05 | not sure to be honest - there are some checks that look at whether a table starts "ii" - but it's not something I've done a lot with | 06:46 |
atrofast | I guess all I can do is try it out then, thanks a lot PaulM05 | 06:46 |
PaulM05 | are the updates a consequence of something else or are they a way to directly update the data? e.g. is it like when you create a sequence it creates a row in iisequences or is that you want to store info in iiblah and the procedure will be an interface for updating that info? | 06:47 |
atrofast | PaulM05: The second, iiblah has info, the procedures can add/remove info from it | 06:50 |
PaulM05 | ok | 06:51 |
gerhard | I have a table with about 4000 rows that contains a field of type ingresdate, this field *should* contain date-only, but a few rows contain date+timestamp. Any way to get rid of the timestamp and just keep the date? | 06:55 |
PaulM05 | I've done it the other way around | 06:58 |
*** cthibert has left #ingres | 07:04 | |
*** cthibert has joined #ingres | 07:07 | |
PaulM05 | how about something like: UPDATE mytable SET date_col=DATE(LEFT(VARCHAR(date_col),11))? (that's based on II_DATE_FORMAT=US) | 07:08 |
withdefault | gerhard, try http://ingres.pastebin.com/AGmaULTY | 07:17 |
withdefault | I get this on my instance http://ingres.pastebin.com/5yEazR7b | 07:22 |
*** Mud has quit IRC | 07:51 | |
gerhard | withdefault: thanks, the date_trunc function does the job. Paul: also thanks, I suppose date(left(... also would have worked. | 07:51 |
PaulM05 | yeah - I wasn't sure if date_trunc would just give you a date with a 00:00:00 time - of course withdefault did the sensible thing and tested it! ;) | 07:53 |
withdefault | I wasn't sure what would happen with times close to boundaries | 07:53 |
*** Mud has joined #ingres | 07:57 | |
*** Mud has quit IRC | 08:07 | |
*** quelgeek has joined #ingres | 08:16 | |
*** gerhard has quit IRC | 08:22 | |
quelgeek | Hi gang. The call for papers for the UK IUA 2010 conference is posted here: http://community.ingres.com/forum/dba-forum/announcement-uk-iua-2010-call-papers-now-open.html | 08:50 |
*** zxiiro has quit IRC | 08:50 | |
quelgeek | The IUA is also asking for nominees for this year's Ingres Community Service award. Please email your nominations to award2010@iua.org.uk | 08:53 |
*** mull has joined #ingres | 09:11 | |
*** Mud has joined #ingres | 09:30 | |
*** PaulM05 has quit IRC | 09:56 | |
*** grantc has quit IRC | 09:56 | |
*** zxiiro has joined #ingres | 10:06 | |
*** ChanServ sets mode: +v zxiiro | 10:06 | |
*** Dejan has quit IRC | 11:00 | |
*** DarylM_iPhone has joined #ingres | 11:22 | |
*** DarylM_iPhone has joined #ingres | 11:24 | |
DarylM_iPhone | Hehe preparing for an iPhone presentation | 11:25 |
*** DarylM_iPhone has quit IRC | 11:26 | |
*** atrofast has quit IRC | 11:29 | |
*** pboro has quit IRC | 11:43 | |
*** pboro has joined #ingres | 11:44 | |
*** Mud has quit IRC | 11:51 | |
*** Mud has joined #ingres | 11:57 | |
*** withdefault has quit IRC | 11:58 | |
*** atrofast has joined #ingres | 12:02 | |
*** KermitTheFragger has quit IRC | 12:02 | |
*** toumi01 has left #ingres | 12:53 | |
*** rossand has quit IRC | 13:08 | |
*** rossand has joined #ingres | 13:15 | |
*** ChanServ sets mode: +o rossand | 13:15 | |
*** rossand has quit IRC | 13:16 | |
*** cthibert has left #ingres | 13:50 | |
*** atrofast has quit IRC | 13:59 | |
*** toumi01 has joined #ingres | 14:15 | |
*** zxiiro has quit IRC | 14:27 | |
*** rossand has joined #ingres | 14:42 | |
*** ChanServ sets mode: +o rossand | 14:42 | |
*** rossand has quit IRC | 14:43 | |
*** rossand1 has joined #ingres | 14:43 | |
*** Mud has quit IRC | 14:51 | |
*** zxiiro has joined #ingres | 15:10 | |
*** ChanServ sets mode: +v zxiiro | 15:10 | |
*** atrofast has joined #ingres | 15:52 | |
*** mull has quit IRC | 15:53 | |
*** atrofast has left #ingres | 17:58 | |
*** cytrinox_ has joined #ingres | 19:20 | |
*** cytrinox has quit IRC | 19:23 | |
*** cytrinox_ is now known as cytrinox | 19:23 | |
*** withdefault has joined #ingres | 23:07 | |
*** ChanServ sets mode: +o withdefault | 23:07 | |
*** Mud has joined #ingres | 23:25 | |
*** Mud has quit IRC | 23:53 | |
*** Mud has joined #ingres | 23:57 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!