| Time |
Nick |
Message |
| 00:04 |
|
alynn26 joined #evergreen |
| 00:25 |
|
plux joined #evergreen |
| 00:36 |
|
alynn26_away joined #evergreen |
| 01:30 |
|
alynn26 joined #evergreen |
| 06:42 |
|
Christineb joined #evergreen |
| 06:59 |
|
agoben joined #evergreen |
| 07:11 |
|
rjackson_isl joined #evergreen |
| 07:12 |
|
collum joined #evergreen |
| 08:03 |
|
alynn26 joined #evergreen |
| 08:26 |
|
mantis1 joined #evergreen |
| 08:38 |
|
mmorgan joined #evergreen |
| 09:01 |
|
sandbergja joined #evergreen |
| 09:09 |
|
Dyrcona joined #evergreen |
| 09:16 |
|
jvwoolf joined #evergreen |
| 09:23 |
|
cmalm joined #evergreen |
| 09:24 |
|
Dyrcona joined #evergreen |
| 09:38 |
|
jvwoolf joined #evergreen |
| 09:58 |
Dyrcona |
Bleh... Struggling with rsync and sudo this morning. It's enough to make me just let the user run any command without a password on the remote. |
| 10:00 |
Bmagic |
@band run any command without a password |
| 10:01 |
Dyrcona |
And, nothing that I find on the Interwebs works.... |
| 10:01 |
Bmagic |
what the heck is the internet? |
| 10:01 |
Dyrcona |
Maybe, I'll just let the user run rsync without a password. |
| 10:02 |
Dyrcona |
Bmagic: I hear it's a series of tubes. |
| 10:02 |
Bmagic |
I heard "pipes" |
| 10:02 |
Bmagic |
Either way, it sucks |
| 10:04 |
Dyrcona |
The Intertubes are actually OK. It's the crap we build on top of it. :P |
| 10:04 |
Bmagic |
Are you saying that Cisco > Linux? |
| 10:04 |
* Dyrcona |
is often amazed that any of this stuff ever works. |
| 10:05 |
Bmagic |
Windows ME didn't work. So there's that. |
| 10:05 |
Dyrcona |
Um, no. I'm saying hardware > software and tcp/ip > higher level protocols. |
| 10:06 |
Dyrcona |
But, the hardware itself is mostly software, now, so its even more broken than usual.... |
| 10:06 |
* Dyrcona |
likes to read all of the CVEs for hardware and software and get depressed. It's enough to make me want to raise goats and live in a hut. |
| 10:07 |
Dyrcona |
The last CVE ever filed will be: "The AI has decided that humans server no purpose..." |
| 10:07 |
Dyrcona |
Maybe "server" wasn't a typo... :P |
| 10:08 |
Dyrcona |
Anyway, back to messing rsync and sudo to put /etc/hosts in place without having to actually login as a user on the server. |
| 10:11 |
Dyrcona |
Fun thing: When I try one of the options that "works" according to the stackoverfolow network, I get a protocol mismatch error and messages asking if the shell is clean... So, sudoers file, it is. |
| 10:12 |
Bmagic |
You go with your bad self |
| 10:13 |
Bmagic |
Show that file who's boss |
| 10:13 |
|
nfBurton joined #evergreen |
| 10:13 |
Dyrcona |
Well, it's risky cause now files can be copied anywhere on the server, but you have to have a key for access, and you have to ssh tunnel, and the tunnel host also requires a key....so, it comes down to whether or not I trust the other staff that have access to the machines. |
| 10:14 |
Bmagic |
jail em. or make a Docker image that can be reset. or something else clever |
| 10:14 |
nfBurton |
Anyone know how to fix the Ejabberd Could not open TCP socket to Jabber Server: Connection Refused error when trying to start osrf services |
| 10:14 |
Dyrcona |
And, they can still put files anywhere. It's just a multi-step process. |
| 10:15 |
Bmagic |
nfBurton: classic |
| 10:15 |
nfBurton |
lol |
| 10:15 |
Dyrcona |
Bmagic: A docker image for a VM host? That'll work well. |
| 10:15 |
nfBurton |
I restarted my server and I cant start the services now |
| 10:15 |
Dyrcona |
nfBurton: It was working before? |
| 10:15 |
Bmagic |
Dyrcona: Docker contain inside* the VM. Do it all the time |
| 10:15 |
nfBurton |
It got tempermental while on vacation |
| 10:15 |
nfBurton |
Yup |
| 10:16 |
Dyrcona |
Bmagic: I'm trying to update files on the hosts all at the same time, so Docker's not gonna solve my problem. :) |
| 10:16 |
Dyrcona |
nfBurton: disk space ok? |
| 10:16 |
Bmagic |
sudo rm -Rf / |
| 10:16 |
jeff |
nfBurton: check if ejabberd is running, and if not, attempt to start? if it fails, it might be due to the hostname changing. the default debian/ubuntu config for ERLANG_NODE is sensitive to hostname changes. |
| 10:17 |
nfBurton |
ejabberd starts but I cant connect |
| 10:17 |
nfBurton |
My disc is at 100 though it seems |
| 10:18 |
Bmagic |
nfBurton: 100% disk full? That'd do it |
| 10:18 |
Bmagic |
handy command for completely removing ejabberd whilst retaining the config: updatedb && cp /etc/ejabberd/ejabberd.yml ~/ && apt-get -y --purge remove ejabberd && locate ejabberd|xargs rm -Rf |
| 10:19 |
jeff |
if you have a full filesystem, i'd resolve that first. first suspect would be large logfiles in /var/log or /openils/var/log |
| 10:19 |
Dyrcona |
So, I'll just do what I do on the VMs: Let this user run any command with NOPASSWD. What can possibly go wrong? :P |
| 10:20 |
Dyrcona |
yeah, what jeff said. I was gonna say "DELETE ALL THE LOGS!!!" |
| 10:20 |
Bmagic |
echo '' > /opensrf/var/log/osrfsys.log - one of my fav's |
| 10:21 |
jeff |
Bmagic: I'm not sure I'd describe that command the way you did. It makes a copy of the config file, but otherwise purges the ejabberd config, user database, etc. Also, it will potentially be far more destructive than described. |
| 10:21 |
Bmagic |
jeff: seems to be working so far :) - but you're right, use with caution |
| 10:22 |
Dyrcona |
I use "truncate -s0 /openils/var/log/*.log" but not so much anymore, since I configure the use of syslog and that leads to autorotation of the logs. |
| 10:23 |
Dyrcona |
Bmagic: BTW, ": > file" does the same thing. |
| 10:23 |
Bmagic |
Dyrcona: Never seen that one. The "colon" command? sweet |
| 10:23 |
Dyrcona |
I use that when "truncate -s0 .." is too much to type. |
| 10:24 |
Dyrcona |
: evaluates to "true" in sh and derivatives, i.e. 0. |
| 10:24 |
Bmagic |
cool |
| 10:24 |
Dyrcona |
I forget exactly why, but I had to do something like " : && command...." in a script once. |
| 10:25 |
Bmagic |
Neato. Reminds me of the old command for Windows that makes no sense: "cons/cons" |
| 10:25 |
Bmagic |
you type that command into the run prompt on Windows 95/98: instant blue screen |
| 10:26 |
Dyrcona |
I know I've used it to force an executable file to open a particular program to run the rest of the "script." |
| 10:27 |
nfBurton |
Thanks for that, that helped. |
| 10:27 |
nfBurton |
Bmagic++ |
| 10:27 |
Bmagic |
the obvious connection from "cons" is "consoto" which is MS's "generic" naming sort of like "Consortium" shortname="CONS" |
| 10:29 |
Dyrcona |
Bmagic: Yeah, if the first line of a shell script begins with ":;exec command ..." Then the shell instance spawned to run it gets replaced with command. |
| 10:30 |
Bmagic |
pretty cool |
| 10:31 |
|
sandbergja joined #evergreen |
| 10:31 |
Dyrcona |
It's usually not necessary, but I use it for ... Emacs batch scripts. So I can write scripts in Emacs Lisp and run them from the command line. :) |
| 10:32 |
Dyrcona |
Anyway, back to fixing my permissions issues on the VM hosts. |
| 10:32 |
Dyrcona |
nfBurton: Glad it was that simple. |
| 10:32 |
Bmagic |
Good luck |
| 10:33 |
Dyrcona |
Bmagic: This is easy, just take a sudoers file I have for the opensrf user and change the username, then copy it to the servers, and put it in place. :) |
| 10:33 |
Bmagic |
Sounds like a job for..... Dyrcona! |
| 10:34 |
Dyrcona |
Definitely not an emacs batch script. :) |
| 10:36 |
Dyrcona |
Oh, one last thing about :;exec, it lets you bypass the shell's limit on number of arguments on the shebang line. |
| 10:36 |
Bmagic |
Muy bien |
| 10:37 |
Dyrcona |
That's the main reason that I use it for my Emacs batch scripts, which, honestly, I almost never create. |
| 11:13 |
|
mllewellyn joined #evergreen |
| 11:13 |
|
mllewellyn left #evergreen |
| 11:14 |
Dyrcona |
Oh, Launchpad! First, you make me log in again, and then you tell me to try again later when adding a comment to a bug. |
| 11:41 |
|
jvwoolf joined #evergreen |
| 11:49 |
|
jihpringle joined #evergreen |
| 12:14 |
|
alynn26 joined #evergreen |
| 12:25 |
|
khuckins joined #evergreen |
| 12:33 |
Dyrcona |
Heh. Having set bug 1850156 to confirmed, I realize that there's only 1 target I need to run because I usually install the other required packages anyway, so I could get by with out a *-git install target. |
| 12:33 |
pinesol |
Launchpad bug 1850156 in Evergreen "Need a -tester make target" [Wishlist,Confirmed] https://launchpad.net/bugs/1850156 |
| 12:36 |
Dyrcona |
Think I'll give others some time to comment before I actually start a branch. Basically all I need is make -f Open-ILS/src/extras/install/Makefile.common install_node_js_from_source. So, I suppose an alternative would be to document that in the README for sites installing from a git checkout. |
| 12:50 |
mmorgan |
I'm trying to add a call number to the autorenewnotify action trigger. I've added target_copy.call_number to the environment, but as soon as I do, I get this error: |
| 12:50 |
mmorgan |
Can't call method "call_number" on an undefined value at /usr/local/share/perl/5.24.1/OpenILS/Application/Trigger/Reactor.pm line 129. |
| 12:50 |
mmorgan |
What am I missing? |
| 12:51 |
|
collum joined #evergreen |
| 12:53 |
Dyrcona |
Oh, lovely.... Run install_nodejs_from_source by itself and it fails because the URL is set in a variable that isn't set in Makefile.common. |
| 12:54 |
Dyrcona |
mmorgan: I thought call number was already available in the environment. |
| 12:57 |
|
collum_ joined #evergreen |
| 12:59 |
mmorgan |
Dyrcona: not seeing it, unless I removed it, but don't think so. Checking. |
| 13:00 |
Dyrcona |
mmorgan: We have "target_copy.call_number.record.mattrs" in our environment for autorenewnotify with no issues. That should make all call_number fields available. You'd also have to add the prefix and suffix if you wanted the labels for those. |
| 13:01 |
Dyrcona |
Based on the id number of the environment entry, I'd say we added that path, and it isn't stock. |
| 13:01 |
* mmorgan |
will try it |
| 13:02 |
Dyrcona |
Do you have to restart the trigger service after changing something like that? |
| 13:03 |
Dyrcona |
I can never remember what requires restarts and what doesn't. |
| 13:04 |
Dyrcona |
target_copy.call_number by itself, should work. |
| 13:04 |
mmorgan |
Maybe the restart was what I was missing? |
| 13:05 |
|
sandbergja joined #evergreen |
| 13:08 |
mmorgan |
still doesn't work after restart :-( |
| 13:10 |
* mmorgan |
will need to revisit after lunch. |
| 13:30 |
|
collum joined #evergreen |
| 14:01 |
jeff |
mmorgan: are you calling get_copy_bib_basics in your template and passing it an object that used to be a raw id, and now needs to be foo.id? |
| 14:02 |
jeff |
mmorgan: looks like the stock template for autorenew does that: |
| 14:02 |
jeff |
[%- SET cid = circ.target_copy || udata.copy -%] |
| 14:02 |
jeff |
[%- SET copy_details = helpers.get_copy_bib_basics(cid) -%] |
| 14:03 |
jeff |
if by adding target_copy.call_number to the environment you've caused circ.target_copy to be a fleshed object and not an ID, you'll want to change SET cid = circ.target_copy to SET cid = circ.target_copy.id and try again |
| 14:06 |
Dyrcona |
jeff++ |
| 14:06 |
Dyrcona |
I think you've got it. |
| 14:07 |
jeff |
not sure if the udata.copy bit would need any modification. looking at commit a9c3660f2b7 would probably be a starting point. |
| 14:07 |
pinesol |
jeff: [evergreen|Cesar Velez] LP#1779920 - Autorenew Feature - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=a9c3660> |
| 14:07 |
|
Stompro joined #evergreen |
| 14:08 |
jeff |
or just reviewing the user_data bits of Open-ILS/src/perlmods/lib/OpenILS/Application/Trigger/Reactor/Circ/AutoRenew.pm |
| 14:09 |
jeff |
looks like udata.copy will be fine, and will already be an id, not an object. |
| 14:11 |
jeff |
I wonder though if there are any cases where circ.target_copy will not be an object, since the SET cid seems to accommodate the possibility that it will not be set. |
| 14:11 |
|
jvwoolf1 joined #evergreen |
| 14:13 |
jeff |
circ is the core object for the autorenewal hook, and target_copy on circ is NOT NULL, so... |
| 14:13 |
Dyrcona |
Yeah. Not sure what's going on. |
| 14:14 |
jeff |
...I'm not sure when you'd ever have an AutorenewNotify where [%- SET cid = circ.target_copy || udata.copy -%] would fall through to using the value in udata.copy. |
| 14:16 |
Dyrcona |
Our template is customized. I'd have to dump it to see if I handled that, but I must have because it is working for us. |
| 14:17 |
jeff |
interesting: "all autorenewals occur from patron's Home OU" |
| 14:17 |
jeff |
I would want them to behave like OPAC renewals. |
| 14:22 |
jeff |
looks like even though that comment exists in OpenILS::Application::Trigger::Reactor::Circ::AutoRenew, the renewal might still respect the value of circ.opac_renewal.use_original_circ_lib |
| 14:25 |
Dyrcona |
Yeah, we've not had any complaints on that front. |
| 14:37 |
jeff |
doesn't look like any stock templates have an example of handling unfleshed-or-fleshed, but that might be a useful addition to the helper, or several helpers. |
| 14:44 |
Dyrcona |
mmorgan | jeff: From our template: copy_details = helpers.get_copy_bib_basics(circ.target_copy.id) |
| 14:45 |
Dyrcona |
So, yeah, I ajusted it for target_copy being fleshed. |
| 14:46 |
Dyrcona |
However, I *think* the helpers work with an object or an id? I'll have to check the code to be certain. |
| 14:46 |
|
rfrasur joined #evergreen |
| 14:48 |
Dyrcona |
Ah, whaddya know... That was the wrong template, 'cause I was looking at an update script with multiple templates in it. :) |
| 14:48 |
jeff |
heh. |
| 14:49 |
jeff |
current master get_copy_bib_basics helper does not seem to handle being fed an object. |
| 14:49 |
Dyrcona |
Huh. Wonder why we're fleshing all the way to record.mattrs. We're not using any of the mattrs. |
| 14:52 |
* mmorgan |
has a meeting to go to now, will try those suggestions a bit later and report back. |
| 14:52 |
mmorgan |
Dyrcona++ |
| 14:52 |
mmorgan |
jeff++ |
| 14:53 |
jeff |
amusingly, get_copy_price helper variables imply that you should pass it an id, but it'll fail unless you hand it an object. :P |
| 14:53 |
jeff |
(every stock template that uses get_copy_price hands it an object) |
| 14:54 |
Dyrcona |
This is from our actual template in the database: [% SET cid = circ.target_copy.id || udata.copy %] |
| 14:54 |
Dyrcona |
The one in that update script was wrong. :) |
| 14:54 |
jeff |
yeah, looks like you modified from stock when you changed the environment. |
| 14:54 |
Dyrcona |
Even when I scrolled down the autorenewnotify update. |
| 14:55 |
Dyrcona |
Yeah, definitely. |
| 14:55 |
Dyrcona |
I was sure we had. |
| 14:55 |
Dyrcona |
And we are using the attrs in the template in the database. |
| 14:56 |
Dyrcona |
We're using icon_format to fill in the atrr.value as a "format" field. |
| 14:57 |
Dyrcona |
I could pastebin the template if anyone is really interested. |
| 14:57 |
* mmorgan |
raises hand |
| 14:57 |
Dyrcona |
:) |
| 14:57 |
mmorgan |
We've thought of adding format, too. |
| 15:01 |
|
mmorgan1 joined #evergreen |
| 15:02 |
Dyrcona |
mmorgan1: https://pastebin.com/yuQmqXkn |
| 15:03 |
|
khuckins joined #evergreen |
| 15:11 |
|
jvwoolf joined #evergreen |
| 15:25 |
|
mmorgan joined #evergreen |
| 15:29 |
mmorgan |
Dyrcona++ |
| 15:37 |
|
mantis1 left #evergreen |
| 16:23 |
mmorgan |
jeff++ |
| 16:23 |
mmorgan |
Making the change to SET cid did the trick |
| 16:23 |
mmorgan |
[%- SET cid = circ.target_copy.id || udata.copy -%] |
| 16:29 |
Dyrcona |
mmorgan: We also added circ.target_copy.parts.part to the environment. |
| 16:31 |
mmorgan |
Yes, did that, too. Whenever we use a call number, we always include its friends, Prefix, Suffix, and Part :) |
| 16:37 |
Dyrcona |
Ah, ok, cool. |
| 17:05 |
|
mmorgan left #evergreen |
| 18:16 |
|
jihpringle joined #evergreen |
| 19:10 |
|
cmalm joined #evergreen |
| 21:37 |
|
sandbergja joined #evergreen |
