The autorefresh identities list is built from unify output in a SortingHat state.
If we change t...

All the deployment is done from packages pip created using the git repositories
defined in requi...

Remove 'unified_releases' directory as this release method
has been moved to https://github.com/...

This patch fixes the composition of params for perceval backends.
The param 'fetch-cache' is rep...

This is the first step towards the transition to a docker
image based on pip packages for Grimoi...

grimoire-kidash is now kidash, and it is unbundled from grimoire-elk.
This patch should make mor...

In redis the queue with raw items could be huge. In order to avoid issues
with the python memory...

In order to avoid that during the checking of the memory size for the arhut python dict
the memo...

To avoid feeding items from redis at a higher speed that we can process them
don't feed items if...

Only upload (import) panels (dashboards) corresponding to the
data sources in the mordred config...

https://github.com/chaoss/grimoirelab

Now the projects are not loaded anymore during config so they
must be loaded before doing the co...

The option skip_initial_load has been removed because it is
not usable anymore.

Now the project...

Start collect phase only if there are items for a repository.

If in mordred config file the option:

strict_mapping = false

is defined, then --no-email-valid...

The compute of the memory usage by the redis raw items dictionary is
pretty expensive once it is...

GrimoireELK collect the raw items directly from Redis.

The autorefresh identities list is built from unify output in a SortingHat state.
If we change t...

Add Arthur support to mordred

In the current implementation, all the raw items for all data sources
in mordred are collected t...

The initial checks for Elasticsearch and Arthur are done
now completly inside Mordred and if the...

Following this approach, the raw items are stored in the same raw indexes
we are using before ar...

RuntimeError: Lost connection to MySQL server during query (err: 2013)

To avoid this error I ha...

Read https://caremad.io/posts/2013/07/setup-vs-requirement/ for the
details about that. So we do...

Install httpprety pip also.

With the sortinghat load command a param --reset could be provided
in order to unmerge all ident...

Merges #42
Closes #42

If this strict-mapping option is set to False during unify the matching
will be done using a no ...

With this approach, we avoid having deadlocks

Now the releases specification files are published in
https://raw.githubusercontent.com/grimoire...

[task_identities] Fix downloading of identities file from GitHub

If the studies are executed just after the enrich process finishes,
some items are not yet index...

[panels] Fix all panels issues

Present the error causing a RuntimeError while reading configuration
in a slightly more friendly...

Mordred now supports the loading of all panels and the aliases creation for them.

Now -c accepts a list of configuration files, instead of
only one. If only one is specified, it ...

The queues used to share the SH uuids and the data sources to refresh must
be cleaned after the ...

It is not needed anymore that the organizations included in the GrimoireLab
yaml identities file...

Until now we use a GitHub raw internal URL which included a token.
This URL won't work anymore a...

tasks are not done during enrichment tasks execution.

local files.

Added a new bool param affiliate to control if affiliation must be done.
To avoid doing the unif...

from files stored in GitLab.

so we can reuse it in other places.

Using the name parameter, when creating threads, the name of the thread
can be set. These names ...

In ElasticSearc/Kibana 5.6 the .kibana mapping is set to strict,
which means that new items must...