psaiz
psaiz
A full review of the page is needed. There are at least three issues with the current version: 1. It is very LHC centric. There are comments like `Data policies:...
The virtual machines used for elasticsearch clusters for prod/qa/dev are no longer needed. The machines can be destroyed.
Comments from our OpenShift colleagues: ``` $ oc adm top pod NAME CPU(cores) MEMORY(bytes) cache-2-wxtn8 1m 3Mi logstash-deployment-6b5cdbd68f-rl7w9 20m 1569Mi mq-1-txdvz 1m 113Mi proxy-24-cj76v 0m 69Mi proxy-24-ht7xc 0m 65Mi web-16-krrxw...
The schema of opendata defines 3 objects (records, glossary and docs). All these obejcts are stored under indices like `record-`. The searches are done in all the indices. There are...
There should be documentation explaining how the users can filter results, and the syntax that it is accepted in the query filter. For instance, things like `(-"path information") High-Level Trigger`...
The service should appear [here](https://monit-grafana.cern.ch/d/000000855/service-availability?orgId=1). In order to do that, we have to follow the [monit procedure](https://monit-docs.web.cern.ch/metrics/http/#service-availability): * Identify how we are going to calculate our availability * Send the...
Currently, the eos monitoring is done at the file level. It would be nice to have statistics also at the dataset level
We collect statistics based on the nginx access monitoring. The information is available [here]( https://monit-timberprivate.cern.ch/dashboards/goto/ff3f66e8503c4aefed9be859f1d2d2fe?security_tenant=opendata) The accesses can be split into categories, which are done based on invenio [here] (https://github.com/inveniosoftware/helm-invenio/blob/master/invenio/templates/configurations/logstash.yaml#L112)...
Currently, COD uses the statistics provided by EOS to see the usage of the last 30 days. See https://monit-timber.cern.ch/dashboards/goto/9654e31aa03266c33431f5fa4c50b74c?security_tenant=opendata It would be nice to have long-term data, going over the...
Currently, the COD portal is deployed in openshift following the instructions from https://gitlab.cern.ch/cernopendata/deployment. It might be better to use the helm charts provided by our invenio colleagues https://github.com/inveniosoftware/helm-invenio