James Allen
3d195de3e9
Improve testing of archiving and be explicit about include_deleted parameter
2016-12-05 16:48:01 +00:00
James Allen
7717625534
Include deleted docs when archiving
2016-12-05 16:31:51 +00:00
James Allen
bd5086c054
Don't return deleted docs with all project docs
2016-12-05 14:37:44 +00:00
James Allen
e156d8017c
Be explicit in version filter
2016-11-30 15:29:43 +00:00
James Allen
8b95a04a36
Don't look up version in mongo if not needed
2016-11-30 15:08:56 +00:00
James Allen
50f2983912
Log out version when updating doc
2016-11-29 16:49:35 +00:00
James Allen
eb9a3b314a
Set/get version in docOps via docstore HTTP API
2016-11-28 14:55:16 +00:00
Brian Gough
7a28fb58e7
clean up mongo query
2016-09-05 11:14:41 +01:00
Brian Gough
0a0a8e23fe
remove unused modules
2016-09-05 11:14:41 +01:00
Henry Oswald
1541eb319c
add more null checks to health check
2015-12-02 10:10:52 +00:00
Henry Oswald
a964119b5e
check err in health check and null check res
2015-12-01 21:50:18 +00:00
Henry Oswald
29c812cece
moved the build opts down
2015-10-16 12:57:05 +01:00
Henry Oswald
4f5669591a
add timeout to health check requests and always try the delete
2015-10-16 12:30:53 +01:00
Henry Oswald
de9a24f7bb
pulled health check into its own file and use project_id from settings
2015-10-16 10:13:54 +01:00
Henry Oswald
f1bf0c97ec
make archive and unarchive parallelLimit to speed it up
2015-09-22 09:41:26 +01:00
Henry Oswald
9cfa4b3f84
Don't put docs which are already in s3 through archive job
...
This fixes the bug we saw with 'RangeError: Maximum call stack size exceeded'
if lots of docs are already in s3 cb() gets called synchronously multiple times quickly
which can cause the execption. I am not sure where the recursion is, maybe inside async.
doing setImmediate(cb) also fixes the issue as I beilve it gives the
process chance to clear the stack. Similar to process.nextTick
2015-09-07 14:11:02 +01:00
Henry Oswald
738e33a052
change crypto to always use utf-8 as it works with aws
2015-09-01 16:52:02 +01:00
Henry Oswald
599b9770f4
add un archive endpoint
2015-08-13 22:48:45 +01:00
Henry Oswald
ef4e437c3f
improve doc manager so it calls get doc recursivly
2015-08-13 22:47:33 +01:00
Henry Oswald
b8884352f6
added some better error logging
2015-08-13 13:59:22 +01:00
Henry Oswald
859e8a0e57
- deal with calling archive twice on same project
...
- clean up unit tests
- change s3 keys
- use err not error
2015-08-13 13:00:09 +01:00
Henrique Dias
42f2afc4dd
remove unarchived docs from s3
2015-06-03 20:05:15 -03:00
Henrique Dias
aa23c1da56
md5 check for doc lines
2015-06-03 11:45:47 -03:00
Henrique Dias
5ba14522da
handling error
2015-06-02 20:08:50 -03:00
Henrique Dias
a26320013d
more acceptance tests
2015-06-02 19:24:45 -03:00
Henrique Dias
0cd2120430
more unit tests for archive feature
2015-06-02 18:13:16 -03:00
Henrique Dias
5c31e80062
more refactoring
2015-06-02 16:29:32 -03:00
Henrique Dias
ddd26798f1
refactoring
...
s3 filter in router
2015-06-02 15:55:22 -03:00
Henrique Dias
86f16caeef
fix some issues
2015-06-02 15:13:28 -03:00
Henry Oswald
ea92f57acc
point where we unset doc lines also search on doc rev to make sure
...
the document has not been updated
2015-06-02 12:39:49 +01:00
Henry Oswald
60bc7bf4e1
pulled logic for arching docs arpart a bit, can be moved into its own file
...
hacked out (needs tests) the unarchiving of individual docs & doc updates
2015-06-02 12:36:38 +01:00
Henry Oswald
19f1a9ad82
changed style of async jobs to map to use underscore which is like we
...
use in the rest of the stack. Also removed a few => binds which are
not needed
2015-06-02 10:59:35 +01:00
Henrique Dias
7c7ff64904
retrieving all docs from s3
2015-06-01 19:36:26 -03:00
Henrique Dias
7f5204f0f7
initial version for archive docs in S3
2015-06-01 18:24:40 -03:00
Henry Oswald
d59fb0d590
Merge branch 'master' into doc-collection-only
...
Conflicts:
app/coffee/DocManager.coffee
2015-02-27 14:58:38 +00:00
Henry Oswald
53480153e3
if the doc has no lines then it is new and needs to be written
2015-02-27 14:50:08 +00:00
Henry Oswald
32829f0e6e
remove old rev, not needed any more
2015-02-27 14:06:06 +00:00
Henry Oswald
a73dc90b00
change mongomangers upsert to use an inc
...
this is tested and does work with new documents
2015-02-26 16:01:10 +00:00
Henry Oswald
7242119532
simplified DocManager.updateDoc so it calls mongomanager.findDoc
...
makes the logic for dealing new documents that are not in mongo yet
much simpler
2015-02-26 16:00:28 +00:00
Henry Oswald
5861231548
docstore now only touches doc collection
...
- need to run migration 1 before using this for this to work
- added acceptence test so new documents updated should be upserted, not return 404
2015-02-26 14:54:08 +00:00
Henry Oswald
3cd3c118d1
don't log out when doc lines are [] as this is a new empty doc
2015-02-23 13:17:55 +00:00
Henry Oswald
daf7c4f68e
log out when a doc from the project collection is used
2015-02-22 18:45:21 +00:00
Henry Oswald
a0832c79e6
return 404 if doc is deleted and include_deleted is not set to true
2015-02-03 14:05:08 +00:00
Henry Oswald
7b0b72494f
write to doc collection first
2015-01-27 15:05:35 +00:00
Henry Oswald
5d2c09bf75
read docs from doc collection first
...
- removes include deleted flag as it is irrelivant now
2015-01-22 15:05:48 +00:00
Henry Oswald
535d4435bf
added missing err check
2015-01-21 12:31:06 +00:00
Henry Oswald
704bf53fe5
Write docs to do collection and project collection
...
- docs are written to project collection then doc collection
- deletes used to insert into doc collection, now its the same upsert
with deleted = true
- does not set delted = false for new docs as it is not needed, save space
- changed delete to be seperate update on mongo
2015-01-21 11:52:39 +00:00
Henry Oswald
28990d5a39
changed logging to easier to see when the http request starts
2015-01-20 13:46:57 +00:00
James Allen
6d24c69e96
Don't return deleted docs by default, make it a flag
2014-06-06 12:37:42 +01:00
James Allen
4818ae8131
Use flag to only get deleted doc when appropriate
2014-06-05 14:10:44 +01:00
