| # | Time | Nick | Message |
|---|
| # | 06:43:29 | phasefx has quit IRC |
| # | 06:53:05 | phasefx has joined #evergreen |
| # | 07:05:21 | Dyrcona has joined #evergreen |
| # | 07:20:41 | Dyrcona has a hate on for his legacy ILS. |
| # | 07:57:24 | sfortin has joined #evergreen |
| # | 08:43:07 | tsbere has joined #evergreen |
| # | 08:50:41 | gdunbar has joined #evergreen |
| # | 09:00:48 | alxp has joined #evergreen |
| # | 09:06:09 | Meliss has joined #evergreen |
| # | 09:09:14 | rjackson-isl has joined #evergreen |
| # | 09:17:27 | shopkins has joined #evergreen |
| # | 09:21:33 | jenny has joined #evergreen |
| # | 09:28:27 | tsbere | phasefx, on your latest commit I noticed (just after your changes) a typo in a comment. "pull old car off the cards list" |
| # | 09:29:33 | phasefx | tsbere: I'll get it, gracias |
| # | 09:33:43 | kmlussier has joined #evergreen |
| # | 10:22:51 | miker_g1 has joined #evergreen |
| # | 10:28:06 | bshum has joined #evergreen |
| # | 10:41:11 | youdonotexist has joined #evergreen |
| # | 10:45:27 | yboston has joined #evergreen |
| # | 10:52:52 | kmlussier has quit IRC |
| # | 10:55:15 | kmlussier has joined #evergreen |
| # | 11:06:38 | toubs has joined #evergreen |
| # | 11:06:54 | collum has joined #evergreen |
| # | 11:07:41 | toubs | Hi all. Is there anybody that can help with installing evergreen on centos? |
| # | 11:08:34 | jeff | toubs: i can try. others have more experience in that area, but -- what's your question? |
| # | 11:08:59 | miker_ | toubs: depends on the version. trunk has RHEL/centos/fedora support in the pre-reqs makefile now, IIRC |
| # | 11:09:47 | toubs | I am running centos 5.4 |
| # | 11:11:15 | toubs | I have installed the pre-reqs but when I try to install Opensrf I am getting a problem when running the "make -f src/extras/Makefile.install" |
| # | 11:11:26 | toubs | The error I am getting is |
| # | 11:11:44 | toubs | syntax error at -e line 1, near ""RPC::XML";" |
| # | 11:11:51 | toubs | Execution of -e aborted due to compilation errors. |
| # | 11:12:01 | toubs | make: *** [install_centos_perl] Error 9 |
| # | 11:12:21 | toubs | I am not sure what package of perl it is trying to load. |
| # | 11:13:27 | berick | perhaps Makefile.install is broken. there have been a few changes to file recently, some errors could have slipped in |
| # | 11:13:59 | afterl has joined #evergreen |
| # | 11:14:37 | jeff | toubs: you say this is with the OpenSRF Makefile.install -- what version of OpenSRF are you working with? |
| # | 11:15:02 | toubs | OpenSRF-1.2.2 |
| # | 11:17:36 | jeff | yes, there's a bug in the makefile. |
| # | 11:18:38 | toubs | Thnak You. - I have been trying to figure this out for the last 2 days now at least I now know I am not doing anything wrong. |
| # | 11:20:50 | toubs | Where will I be able to find an older version that will work? |
| # | 11:21:54 | jeff | on line 352, there's a call to install modules without testing them (because they have broken tests). it seems to at the very least have an unbalanced parenthesis. |
| # | 11:22:25 | jeff | berick, miker_, are either of you familiar with the notest pragma to CPAN to fix that? |
| # | 11:24:06 | miker_ | jeff: I'm not ... it fails even when downloaded and the installation is attempted by hand? assuming so, I'd have no problem with sidestepping their tests -- though reporting upstream would be a GoodThing(tm) if someone were to do that |
| # | 11:24:29 | jeff | for m in $(CENTOS_PERL_NOTEST); do perl -MCPAN -e "CPAN::Shell->notest('install', \"$$m\";"; done |
| # | 11:24:48 | berick | looks liek the second-to-last ; needs to be a paren (what you said jeff) |
| # | 11:25:08 | jeff tries |
| # | 11:25:09 | atz | the package maintainer should already be aware since perltesters will have noted the failure |
| # | 11:25:31 | atz | if it is really just failing tests on a given distro |
| # | 11:25:53 | atz | still good to file a bug report, but it shouldn't surprise them |
| # | 11:28:02 | berick | and now for something completely different.. sip2 offers a variable length "item properties" field (CH) in the Item Info response. It's currently unused in the Everegreen SIP integration code. I can only find 1 reference to it online, which is an OCLC doc talking about how you could use to pass along "size, weight, etc.". |
| # | 11:28:41 | berick | anyone familiar with this field? would this be an a good place to put arbitrary copy attributes (like whether the copy is floating, etc.)? |
| # | 11:29:17 | berick | and, if so, we'll need some kind of structure for the values |
| # | 11:29:48 | berick | atz: do you recall if koha uses this field? |
| # | 11:29:53 | atz | spec says "It may possibly be used for security reasons" |
| # | 11:30:03 | atz | (not much help there) |
| # | 11:30:35 | atz | checking koha |
| # | 11:30:36 | berick imagines that sentence ending "...or not" ;) |
| # | 11:30:54 | jeff | security as in, "does the weight of the items on the pad exceed the cataloged weight of the book" :) |
| # | 11:31:43 | berick | "is the book too big to fit through the door" |
| # | 11:32:06 | atz | jeff: yeah, that's the idea, i'm guessing... |
| # | 11:32:31 | miker_g1 has quit IRC |
| # | 11:33:58 | atz | koha doesn't seem to do anything w/ it |
| # | 11:34:07 | miker_g1 has joined #evergreen |
| # | 11:34:13 | jeff | toubs: your best bet at the moment would be to comment out the following lines in the Makefile.install, then install that module by hand first: |
| # | 11:34:16 | jeff | 142 CENTOS_PERL_NOTEST = \ |
| # | 11:34:18 | jeff | 143 RPC::XML |
| # | 11:34:26 | berick | atz: thanks for checking |
| # | 11:34:27 | atz | builds it into messages if it is populated, but never populates it w/ anything |
| # | 11:35:19 | atz | it gets sent (optionally) on both checkin and checkout and both their responses |
| # | 11:35:35 | jeff | toubs: by commenting out the definition of CENTOS_PERL_NOTEST, you'll skip the buggy attempt to install it in the Makefile. you'll want to install it by hand using something like this: "perl -MCPAN -e shell" then type "install RPC::XML" |
| # | 11:36:08 | jeff | toubs: it may fail. if it does, collect some output from the failed attempt and ask here -- you can put the output in the pastebin at http://paste.lisp.org/new/evergreen |
| # | 11:49:35 | toubs | thanks guys |
| # | 11:49:46 | toubs | I will give it a try and let you know |
| # | 11:50:01 | jeff | toubs: good luck! thanks for reporting the trouble you encountered! |
| # | 11:50:33 | shopkins has quit IRC |
| # | 12:00:16 | mrpeters-isl has joined #evergreen |
| # | 12:03:16 | kmlussier has quit IRC |
| # | 12:06:33 | mrpeters-isl | guys, i've got a situation installing OpenSRF 1.2.2 where I cannot get OpenSRF C to start. Is there some diagnostics I can do to determine where the problem is? both Router and Perl are starting up properly, and talking to jabber |
| # | 12:07:46 | jamesrf has joined #evergreen |
| # | 12:12:41 | jeff | mrpeters-isl: did you create /etc/ld.so.conf.d/osrf.conf and run ldconfig? |
| # | 12:12:54 | jeff | mrpeters-isl: are there errors in the logs when you try to start the C services? |
| # | 12:13:01 | mrpeters-isl | jeff...thats a possibility...let me check |
| # | 12:13:09 | mrpeters-isl | ldconfig...that is |
| # | 12:14:16 | mrpeters-isl | jeff: was correct...ran ldconfig again just in case, but still no luck starting c |
| # | 12:14:39 | mrpeters-isl | osrfsys.log:opensrf 2010-08-16 16:14:00 [ERR :16369:osrf_app_session.c:164:] Transport error in recv() osrfsys.log:opensrf 2010-08-16 16:14:00 [ERR :16369:osrf_settings.c:48:] No osrfMessage received from host xenb4apache.evergreen.lib.in.us (timeout?) osrfsys.log:opensrf 2010-08-16 16:14:00 [ERR :16369:osrf_settings.c:67:] Unable to load config for host xenb4apache.evergreen.lib.in.us osrfsys.log:opensrf 2010-08-16 16:14:00 |
| # | 12:15:08 | mrpeters-isl | osrf_settings.c talks to jabber, correct? |
| # | 12:15:15 | mrpeters-isl | to get its settings? |
| # | 12:19:19 | jeff | osrf_settings.c is part of the opensrf C libs. Not an app itself, in this case it's part of one of your C services complaining. |
| # | 12:21:35 | mrpeters-isl | i see |
| # | 12:21:52 | mrpeters-isl | so, where might the disconnect be? i don't see anything incorrect in my opensrf_core.xml |
| # | 12:22:11 | natschil has joined #evergreen |
| # | 12:22:28 | jeff | with srfsh, do you get output for the following command? request opensrf.settings opensrf.settings.host_config.get "localhost" |
| # | 12:22:42 | mrpeters-isl | just one moment |
| # | 12:23:45 | mrpeters-isl | srfsh# request opensrf.settings opensrf.settings.host_config.get "localhost" Unable to communicate with service opensrf.settings |
| # | 12:23:55 | mrpeters-isl | i am, of course, not using localhost however.....so perhaps that last part is incorrect? |
| # | 12:24:13 | mrpeters-isl | same error with my FQDN though |
| # | 12:24:20 | jeff | yeah, that part didn't matter for this test. |
| # | 12:24:26 | mrpeters-isl | ok cool |
| # | 12:24:42 | jeff | you have a problem with the (perl) service opensrf.settings |
| # | 12:24:48 | mrpeters-isl | ok |
| # | 12:25:03 | jeff | check your logs for clues as to why that service failed to start/died |
| # | 12:25:33 | mrpeters-isl | 10-4 |
| # | 12:26:23 | mrpeters-isl | jeff: i'm wondering about this - router 2010-08-16 16:24:08 [DEBG:16495:osrf_router.c:348:] osrfRouterClassHandleBounce() router 2010-08-16 16:24:08 [INFO:16495:osrf_router.c:350:] Received network layer error message from opensrf@private.xenb4jab.evergreen.lib.in.us/opensrf.settings_listener_at_xenb4apache.evergreen.lib.in.us_16505 router 2010-08-16 16:24:08 [WARN:16495:osrf_router.c:357:] We lost the last node in the class |
| # | 12:26:46 | mrpeters-isl | perhaps the problem goes all the way back to the router? any ideas what this "lost the last node" error indicates? |
| # | 12:28:18 | jeff | That warning is because the last node for the opensrf.settings service went away. So far, it looks like the router is doing the right thing. Look earlier in the log to see if opensrf.settings logged anything useful before going away? |
| # | 12:28:25 | miker_g1 has quit IRC |
| # | 12:28:34 | mrpeters-isl | im starting things up again and doing just that |
| # | 12:28:40 | jeff | good luck! |
| # | 12:28:47 | mrpeters-isl | opensrf@xenb4apache:/openils/var/log$ tail opensrf.settings_unix.log 2010/08/16-16:27:37 OpenSRF::UnixServer (type OpenSRF) starting! pid(16589) Binding to UNIX socket file /openils/var/lock/opensrf.settings_unix.sock using SOCK_STREAM Group Not Defined. Defaulting to EGID '1003 1003' User Not Defined. Defaulting to EUID '1003' |
| # | 12:28:54 | mrpeters-isl | seems happy, right? |
| # | 12:29:26 | jeff | anything in osrfsys.log about opensrf.settings? |
| # | 12:29:58 | mrpeters-isl | [2010-08-16 16:27:37] opensrf.settings [INFO:16589:UnixServer.pm:183:] UnixServer: min=5, max=15, min_spare=3 max_spare=5, max_req=1000, log_file=/openils/var/log/opensrf.settings_unix.log, port=/openils/var/lock/opensrf.settings_unix.sock, pid_file=/openils/var/run/opensrf.settings_unix.pid [2010-08-16 16:27:37] /openils/bin/opensrf-perl.pl [INFO:16590:Inbound.pm:98:] loading router info ARRAY(0x1624528) [2010-08-16 16:27:37 |
| # | 12:30:22 | mrpeters-isl | [2010-08-16 16:27:37] /openils/bin/opensrf-perl.pl [INFO:16590:Inbound.pm:113:] opensrf.settings connecting to router router@private.xenb4jab.evergreen.lib.in.us/router [2010-08-16 16:27:37] /openils/bin/opensrf-perl.pl [INFO:16590:Inbound.pm:132:] opensrf.settings inbound: going into listen loop [2010-08-16 16:27:40] opensrf.persist [INFO:16598:UnixServer.pm:183:] UnixServer: min=5, max=25, min_spare=2 max_spare=5, max_req=1 |
| # | 12:30:26 | miker_g1 has joined #evergreen |
| # | 12:30:40 | mrpeters-isl | =/openils/var/lock/opensrf.persist_unix.sock, pid_file=/openils/var/run/opensrf.persist_unix.pid |
| # | 12:30:41 | agJohn | Does data get into action.hold_copy_map via any means other than the hold targeter? Should I expect to find data there to allow "opportunistic" filling of holds (capturing an item upon checkin)? |
| # | 12:30:49 | jeff | so at this point, i'd try the srfsh request again. |
| # | 12:31:08 | mrpeters-isl | still unable to communicate |
| # | 12:31:17 | jeff | agJohn: that is exactly what opportunistic hold capture uses, but it is populated by the hold targeter. |
| # | 12:31:41 | mrpeters-isl | srfsh 2010-08-16 16:31:02 [INFO:16657:osrf_app_session.c:621:] [opensrf.settings] sent 86 bytes of data to router@private.xenb4jab.evereen.lib.in.us/opensrf.settings srfsh 2010-08-16 16:31:02 [DEBG:16657:osrf_app_session.c:624:] Sent: [{"__c":"osrfMessage","__p":{"threadTrace":"0","locale":"en-US","te":"DISCONNECT"}}] |
| # | 12:31:56 | mrpeters-isl | so, it would appear i'm talking to jabber and getting the settings, no? |
| # | 12:31:57 | jeff | agJohn: so no, it isn't entered by other means than the hold targeter, and yes, you should expect to find data there to allow op-capture of holds at checkin. |
| # | 12:31:57 | agJohn | OK. So if I'm getting no captures of opportunistic holds, the hold targeter is not doing its job.... |
| # | 12:32:25 | miker_ | or isn't being run |
| # | 12:32:42 | miker_ | or the hold config is not set up properly and no copies are being found for targetting |
| # | 12:32:50 | jeff | agJohn: that would be the first thing i would check, yes. action.hold_copy_map should have entries, and action.hold_request should have "last targeted" dates (field name escapes me) |
| # | 12:32:59 | jeff | or no copies are in holdable status... |
| # | 12:33:05 | jeff | etc, etc. |
| # | 12:33:12 | miker_ | jeff: prev_check_time IIRC |
| # | 12:33:46 | miker_ | if it's set to "tomorrow at midnight" that means no copies were found |
| # | 12:34:02 | miker_ | er, not tomorrow at midnight, today at midnight |
| # | 12:34:36 | jeff | mrpeters-isl: actually, it looks like your opensrf.settings service sent a disconnect. |
| # | 12:34:53 | jeff | mrpeters-isl: wait, i might have misread. |
| # | 12:34:55 | mrpeters-isl | ah, i thought that was just the end of the request |
| # | 12:36:09 | jeff | mrpeters-isl: in a process list, do you have opensrf.settings processes? say, "pgrep -lf opensrf.settings" |
| # | 12:37:24 | mrpeters-isl | jeff, indeed i do |
| # | 12:37:42 | mrpeters-isl | 4 drones, master, controller, and a listener |
| # | 12:37:49 | mrpeters-isl | sorry...5 drones |
| # | 12:39:29 | brendan_bywater has joined #evergreen |
| # | 12:40:49 | jeff | curious. |
| # | 12:42:13 | miker_ | Sounds like the C services are not using the same hostname as opensrf.xml |
| # | 12:43:23 | miker_ | looks like they're asking for (in the initial error message) settings for xenb4apache.evergreen.lib.in.us, and the settings server says "I KNOW NUZZINK" |
| # | 12:43:43 | jeff | also, testing with srfsh would fail if the hostname in ~/.srfsh.xml was wrong. |
| # | 12:45:47 | jeff | (if it was pointing to the public domain, for example) |
| # | 12:53:15 | atz | i'm getting an error trying to apt-get install pg 8.4 on lenny (w/ backports) |
| # | 12:53:18 | atz | postgresql-server-dev-8.4: Depends: libpq-dev (>= 8.4~) but it is not installable |
| # | 12:54:16 | Dyrcona has quit IRC |
| # | 12:54:19 | atz | not sure why it thinks that isn't installable, so i'm going to attempt that anyway |
| # | 12:57:16 | mrpeters-isl | srfsh is using private |
| # | 12:57:25 | mrpeters-isl | checking the opensrf.xml now |
| # | 12:58:17 | mrpeters-isl | hostname is correct in opensrf.xml |
| # | 12:58:35 | mrpeters-isl | where is the hostname for C services defined? |
| # | 12:58:46 | jeff | how about opensrf_core.xml? |
| # | 12:59:13 | mrpeters-isl | in opensrf_core i simplly replaced my jabber servers |
| # | 12:59:29 | mrpeters-isl | private.localhost becomes private.xenb4jab.evergreen.lib.in.us |
| # | 13:01:24 | miker_ | mrpeters-isl: hostname is devined by the startup script via: perl -MNet::Domain=hostfqdn -e 'print hostfqdn()' |
| # | 13:01:48 | mrpeters-isl | right, which is xenb4apache.evergreen.lib.in.us |
| # | 13:01:52 | jeff | okay, so "grep localhost opensrf_core.xml" should return nothing if those were all replaced. what arguments are you using on osrf_ctl.sh? |
| # | 13:01:55 | miker_ | if not supplied via OSRF_HOSTNAME env var |
| # | 13:02:11 | mrpeters-isl | osrf_ctl.sh -a start_all |
| # | 13:02:30 | jeff | okay, that's good. (not overriding to localhost with -l) |
| # | 13:02:37 | mrpeters-isl | nope |
| # | 13:03:01 | mrpeters-isl | could this be a firewall issue? |
| # | 13:03:40 | jeff | it can almost always be a firewall issue. ;-) |
| # | 13:04:01 | jeff | if you're dealing with a single host, and haven't gone crazy with localhost firewall rules, it's unlikely. |
| # | 13:04:17 | jeff | mrpeters-isl: can you paste your opensrf_core.xml without passwords? |
| # | 13:04:30 | mrpeters-isl | i may not have a zone for this server yet...i just realized |
| # | 13:04:58 | jeff | well, make sure you can resolve your public/private router hostnames |
| # | 13:05:05 | mrpeters-isl | i can |
| # | 13:05:10 | mrpeters-isl | no luck on the firewall either |
| # | 13:05:13 | jeff | good :) |
| # | 13:05:16 | mrpeters-isl | pasting the file for you |
| # | 13:05:18 | jeff | thanks. |
| # | 13:06:06 | lisppaste | mrpeters-isl pasted "opensrf_core.xml" at http://paste.lisp.org/display/113576 |
| # | 13:06:10 | mrpeters-isl | done |
| # | 13:10:05 | jeff | mrpeters-isl: okay, and the <hosts> section at the end of your opensrf.xml file should contain an element that matches xenb4apache.evergreen.lib.in.us -- correct? |
| # | 13:10:26 | mrpeters-isl | indeed it does |
| # | 13:10:47 | mrpeters-isl | <xenb4apache.evergreen.lib.in.us> </xenb4apache.evergreen.lib.in.us> |
| # | 13:11:04 | jeff | mrpeters-isl: is xenb4apache on a different host than xenb4jabber? |
| # | 13:11:33 | mrpeters-isl | it is |
| # | 13:11:51 | mrpeters-isl | however, they have proper hostfile entries (and dns configured) to point to one another |
| # | 13:12:57 | jeff | and xenb4jabber runs only ejabberd? |
| # | 13:13:13 | mrpeters-isl | correct |
| # | 13:13:35 | mrpeters-isl | i have a similar configuration working elsewhere, and i've mirroed it to this brick |
| # | 13:13:46 | rsinger has quit IRC |
| # | 13:19:16 | jenny has quit IRC |
| # | 13:22:01 | tsbere | You accidentally use -l to the osrf_ctl script? |
| # | 13:23:33 | jeff | mrpeters-isl: i'm about out of ideas. what is your {hosts} entry in ejabberd.cfg on xenb4jabber? |
| # | 13:24:11 | rsinger has joined #evergreen |
| # | 13:24:51 | mrpeters-isl | {hosts, ["xenb4jab","xenb4jab.evergreen.lib.in.us","evergreen.lib.in.us", "private.xenb4jab.evergreen.lib.in.us", "public.xenb4jab.evergreen.lib.in.us"]}. |
| # | 13:29:27 | jeff | the important ones are there. |
| # | 13:32:35 | sfortin has quit IRC |
| # | 13:34:59 | phasefx | grabbing 0373 |
| # | 13:35:20 | jeffdavis has left #evergreen |
| # | 13:41:49 | miker_g1 has quit IRC |
| # | 13:45:39 | lisppaste | berick pasted "debian squeeze => "pgsql driver not installed?" WTH" at http://paste.lisp.org/display/113577 |
| # | 13:45:52 | berick | thoughts appreciated ^-- |
| # | 13:48:08 | lisppaste | berick annotated #113577 "autogen.sh output, fwiw" at http://paste.lisp.org/display/113577#1 |
| # | 13:51:50 | natschil has quit IRC |
| # | 13:59:28 | tsbere | Does config.log have any other useful details? |
| # | 14:02:06 | jeffdavis has joined #evergreen |
| # | 14:09:30 | tsbere | berick: Aside from config.log details, perhaps the dbd pgsql driver is the wrong version for libdbi installed? |
| # | 14:14:11 | jeff | berick: from logs, if you have both a system copy and the from-source copy of libdbi installed, you will get that symptom. |
| # | 14:15:00 | berick | hm, ok, will check that |
| # | 14:16:50 | tsbere is trying to figure out why he is listening to Galen tell him how to change basic things like CSS |
| # | 14:18:08 | berick | tsbere: config.log boils down to: | dbi_initialize(0); return(dbi_conn_new("pgsql") ? 0 : 1); |
| # | 14:19:45 | berick | so, ldconfig only finds the 1 copy of libdbdpgsql.so and there's no system packages installed |
| # | 14:19:59 | tsbere | berick: Then I would go with what jeff said, which is kindof an extension of what I said (mismatched versions being found). Did you try ldconfig -p | grep pgsql or dbi? |
| # | 14:20:56 | berick | aha, 2 libdbi.so.0 |
| # | 14:21:21 | berick | there's one in /usr/lib |
| # | 14:21:27 | berick | yep, installed package. |
| # | 14:21:34 | berick wonders what installed that |
| # | 14:21:38 | jeff | aptitude remove libdbi0, re-run ldconfig to be safe. |
| # | 14:21:42 | berick | yeah |
| # | 14:22:01 | berick | pffth, syslog-ng is using it |
| # | 14:22:08 | jeff | using, or recommending? |
| # | 14:22:23 | tsbere | So install the packaged dbd pgsql? |
| # | 14:22:25 | jeff | er, mixing terms. requires, or recommends? |
| # | 14:22:27 | berick | apt-get remove libdbi0 removed syslog-ng as well |
| # | 14:22:32 | jeff | bah. |
| # | 14:23:21 | tsbere | apt-get install libdbi0 syslog-ng libdbd-pgsql then? |
| # | 14:23:28 | jeff | well, it's a known problem with a known fix, but the known fix now breaks something else. |
| # | 14:23:38 | berick | right.. |
| # | 14:23:44 | jeff | tsbere: the packaged version of libdbi in squeeze is old. |
| # | 14:23:56 | jeff | tsbere: and iirc, unsuitable. |
| # | 14:24:04 | tsbere | ok |
| # | 14:24:12 | berick | squeeze wants to install libdbd-pgsql version 0.8.2-1-4.1+b1 |
| # | 14:24:17 | tsbere thinks |
| # | 14:24:27 | berick | yeah, we need 0.8.3 |
| # | 14:24:45 | berick skips syslog-ng for now |
| # | 14:24:48 | jeff | and with squeeze frozen, we either need to determine how to make evergreen work with libdbi in /usr and /usr/local, or we need to find a security bug in 0.8.2 ;-) |
| # | 14:25:55 | berick | or tell syslog-ng libdbi is already isntalled (and where) |
| # | 14:26:24 | jeff | yeah, you could probably force the install, assuming the 0.8.3 from source is suitable for the compiled syslog-ng |
| # | 14:26:40 | jeff | or, last-ditch, install syslog-ng from source afterward. |
| # | 14:28:44 | jeff | looks like the package is maintained in ubuntu, and getting a new maintainer in debian as of about 2 weeks ago: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565562 |
| # | 14:35:53 | berick | until a better solution comes along: |
| # | 14:35:55 | berick | apt-get remove libdbi0; dpkg -i --ignore-depends=libdbi0 /var/cache/apt/archives/syslog-ng_3.1.1-3_i386.deb |
| # | 14:36:02 | berick | ;) |
| # | 14:38:18 | tsbere | apt-get install --nodeps syslog-ng ? |
| # | 14:38:55 | berick | I tried that, actually: E: Command line option --nodeps is not understood |
| # | 14:39:16 | jeff | also, good to specify which dep to ignore, rather than not installing any. |
| # | 14:39:23 | jeff | (dunno if there are any others, in this case) |
| # | 14:46:27 | tsbere | You may want to aptitude hold libdbi0 though |
| # | 14:48:36 | tsbere | Otherwise any update command may put it in because of syslog-ng |
| # | 14:50:19 | berick | good point |
| # | 14:51:51 | berick | heh, debian doesn't like that. it forces you to either install liddbi0 or uninstall syslog-ng |
| # | 14:56:05 | tsbere | install libdbi0, then put it on hold, then manually remove the system copy of the library? |
| # | 15:00:14 | jeff | dirty pool. |
| # | 15:00:58 | tsbere | sometimes the only way to get stuff done is to do it the dirty way, as much as we may not like it. |
| # | 15:01:16 | tsbere | ALT: Remove syslog-ng, then install THAT from source. :P |
| # | 15:15:23 | miker_ | mrpeters-isl: earlier you said "I've mirroed it to this brick" ... how do you mean mirrored? if you mean a disk image, the mnesia db for ejabberd is likely in a broken state and you'll need to apt-get purge and then install it, and then reregister the users |
| # | 15:15:41 | mrpeters-isl | miker_: not mirrored in that sense |
| # | 15:15:55 | mrpeters-isl | by mirrored, i meant i've taken all the same steps, similar hosts files, dns, etc. |
| # | 15:23:12 | mrpeters-isl | appreciate the thought though! |
| # | 15:27:35 | jeff | mrpeters-isl: can you paste the output of "ejabberdctl connected-users" (run on your jabber server) |
| # | 15:28:36 | mrpeters-isl | i can! |
| # | 15:29:52 | mrpeters-isl | jeff: that returns nothing...however, i have no OpenSRF services started...shall i start them? |
| # | 15:30:02 | jeff | ah, yes please. |
| # | 15:30:15 | jeff | i couldn't remember the state things were in last. i should have clarified. :) |
| # | 15:31:11 | lisppaste | mrpeters-isl annotated #113576 "connected users" at http://paste.lisp.org/display/113576#1 |
| # | 15:31:17 | mrpeters-isl | there you are |
| # | 15:32:46 | jeff | mrpeters-isl: have you checked your ejabberd.cfg file to ensure that you've adjusted the max session and other limits and restarted ejabberd? |
| # | 15:33:03 | mrpeters-isl | jeff: i pulled it RIGHT from the other working brick |
| # | 15:33:08 | mrpeters-isl | which is pulled from our production |
| # | 15:33:12 | mrpeters-isl | i just simply changed the hostnames |
| # | 15:33:16 | jeff | interesting. |
| # | 15:34:14 | mrpeters-isl | im going to run a diff on them though, just to check |
| # | 15:34:15 | jeff | i think you should have an active session for the opensrf.settings listener -- you only have some drones. |
| # | 15:34:39 | jeff | mrpeters-isl: i'm most interested in the values specified in step 6 of the opensrf install doc: http://www.open-ils.org/dokuwiki/doku.php?id=opensrf:1.2:install |
| # | 15:34:48 | alxp has quit IRC |
| # | 15:34:50 | mrpeters-isl | right |
| # | 15:35:22 | jeff | i wonder if your max_user_sessions from production is more specific than the generic stock and modified-per-install-docs. |
| # | 15:36:02 | mrpeters-isl | yeah, i have like 5000 user sessions |
| # | 15:36:08 | mrpeters-isl | %% Define the maximum number of time a single user is allowed to connect: {max_user_sessions, 5000}. |
| # | 15:37:07 | jeff | it's possible that your ejabberd.cfg copied from production is in a format that is not suitable for the version of ejabberd your test system here is running. |
| # | 15:37:10 | mrpeters-isl | though, my stanza size is only 1 million, not the 2 million in the documentation |
| # | 15:37:30 | mrpeters-isl | ill bump that to 2 million and see what happens |
| # | 15:38:04 | jeff | your max_user_sessions when pasted in IRC showed on one line, starting with the comment %% -- is isn't commented out on your xenb4jabber system, is it? |
| # | 15:39:17 | mrpeters-isl | no its not just the way it pasted |
| # | 15:40:57 | mrpeters-isl | that did the trick |
| # | 15:41:01 | jeff | recent versions of ejabberd use {access, max_user_sessions, [{10000, all}]}. instead of the old {max_user_sessions, 10000}. syntax. Do you know if you're running ejabberd 1.x in production? |
| # | 15:41:14 | jeff | bumping the stanza size up did the trick? |
| # | 15:41:15 | mrpeters-isl | i dont... |
| # | 15:41:22 | jeff | hooray! |
| # | 15:41:25 | mrpeters-isl | yeah, bumping those to 2 million and C started up |
| # | 15:41:49 | jeff | (it's unclear from the ejabberd docs if the old syntax is supported in 2.x) |
| # | 15:42:11 | jeff | well hooray. could you paste the new output of ejabberdctl connected-users? i'm just curious. |
| # | 15:42:12 | mrpeters-isl | thanks much! will have to consult with Equinox on proper values for our enviornment but at least its up and running now! |
| # | 15:42:15 | mrpeters-isl | i can |
| # | 15:42:32 | lisppaste | mrpeters-isl annotated #113576 "connected v2" at http://paste.lisp.org/display/113576#2 |
| # | 15:43:04 | jeff | yeah, please don't adjust production ejabberd settings based on any of what we've been working on here! ;-) |
| # | 15:43:20 | mrpeters-isl | oh no...this is new hardware...wont go into production like this :) |
| # | 15:43:24 | jeff | looks good! |
| # | 15:44:06 | mrpeters-isl | thanks so much for the help! |
| # | 15:44:25 | jeff | 'welcome! |
| # | 15:47:47 | granitize has quit IRC |
| # | 15:59:34 | Meliss has quit IRC |
| # | 16:06:13 | mrpeters-isl has quit IRC |
| # | 16:18:15 | collum has left #evergreen |
| # | 16:19:33 | bshum has quit IRC |
| # | 16:24:51 | pmplett has joined #evergreen |
| # | 16:29:47 | evette-mlc has joined #evergreen |
| # | 16:42:27 | jamesrf has quit IRC |
| # | 16:55:25 | afterl has quit IRC |
| # | 16:59:26 | tsbere has quit IRC |
| # | 17:00:49 | shadowspar | hmm, in opensrf Makefile.install, libxml-libxml-perl shows up under both the generic_debian and squeeze targets |
| # | 17:00:57 | evette-mlc has quit IRC |
| # | 17:01:09 | shadowspar | presumably an unintended dupe, though I wouldn't think it'd cause any harm |
| # | 17:01:59 | toubs has quit IRC |
| # | 17:10:34 | jenny has joined #evergreen |
| # | 17:14:32 | jenny has left #evergreen |
| # | 17:40:52 | jamesrf has joined #evergreen |
| # | 17:48:46 | Ghidorah has joined #evergreen |
| # | 17:48:55 | Ghidorah | Greetings everyone. |
| # | 17:49:36 | jamesrf | hi |
| # | 17:49:38 | Ghidorah | I have a simple question. I have a TSV and a SQL file to import the data into Evergreen DB. I'm wondering what the syntax is to run the sql file |
| # | 17:51:21 | jamesrf | you can just use the psql client to run an sql script |
| # | 17:51:31 | jamesrf | psql -f /path/to/script.sql |
| # | 17:51:44 | jamesrf | you may need to add parameters for your database host, username, password, etc. |
| # | 17:52:58 | jamesrf | if you are inside a psql shell you can type \i /path/to/file.sql |
| # | 17:54:50 | Ghidorah | Thanks |
| # | 17:55:03 | Ghidorah | Is there any example scripts for bulk user import? |
| # | 17:55:55 | moodaepo | Ghidorah: I don't think so...equinox released some code let me see if I can find the link. |
| # | 18:01:44 | gmcharlt | http://git.esilibrary.com/?p=migration-tools.git;a=summary |
| # | 18:02:49 | yboston has quit IRC |
| # | 18:07:16 | Ghidorah | Any idea on how this tool is used? |
| # | 18:08:37 | gmcharlt | Ghidorah: http://git.esilibrary.com/?p=migration-tools.git;a=blob;f=sql/base/base.sql;h=6d9de88b9a138f1af7965b85fcb83f08aa90c252;hb=HEAD has doc, such as it is |
| # | 18:08:48 | gmcharlt is not claiming that it is particularly comprehensive |
| # | 18:16:17 | yonkeltron has joined #evergreen |
| # | 18:16:58 | yonkeltron has left #evergreen |
| # | 18:29:17 | youdonotexist has quit IRC |
| # | 20:06:24 | brendan_ga has joined #evergreen |
| # | 20:09:43 | brendan_bywater has quit IRC |
| # | 20:44:38 | brendan_bywater has joined #evergreen |
| # | 21:41:23 | jamesrf has quit IRC |
| # | 22:46:32 | jamesrf has joined #evergreen |
| # | 23:13:42 | jeffdavis is now known as jeffd-afk |
April 10-13