This is likely to get worse before it gets better. As part of the Beijing release many components have implemented clustering where one pod becomes three. Clustering is likely to be most common with the data-bases and there are a lot of them.
The OOM team is looking to try to improve at least some of this by trying to eliminate some of the extraneous data-base instances in a three stage process: 1. Use common Helm charts to deploy all of the data-bases in ONAP. Many projects already deploy DBs this way and the OOM team is working with the project teams to convert the rest. 2. Introduce cluster data-base technology into the OOM common charts so that all projects deploy their own DB clusters. Some of this already has been done. 3. Consolidate the multiple instances of DB clusters into shared instances. This is especially important for mariadb (going to mariadb-galera) which has 9 instances in ONAP today. Each of the project teams will need to accept the use of a shared instance so it’s unlikely that we’ll immediately get to a single instance of all DBs. Consolidation of the number of base images has a lot of value with respect to security and general complexity management so there are many reasons to focus on this. There may be many other areas where improvements could be made. Cheers, Roger From: <onap-discuss-boun...@lists.onap.org> on behalf of Andrej Záň <andrej....@pantheon.tech> Date: Friday, April 20, 2018 at 5:39 AM To: Ing-Wher Chen <ing-wher_c...@jabil.com>, "onap-discuss@lists.onap.org" <onap-discuss@lists.onap.org> Subject: Re: [onap-discuss] ONAP's resource requirements, again Hi, I am also interested in this topic. What I have seen so far, many containers are using base images with unnecessary software installed. For example container which does not use node js is based on image which has node js installed (just for illustration) and so on. Are there any plans to optimize it? Regards Andrej Záň ________________________________ Od: Ing-Wher Chen <ing-wher_c...@jabil.com> Odoslané: piatok, 20. apríla 2018 0:00:10 Komu: onap-discuss@lists.onap.org Predmet: [onap-discuss] ONAP's resource requirements, again Hello, I browsed wiki pages and readthedocs, and also some discussion threads on ONAP's hardware resources requirements. Some documentation I found are <https://wiki.onap.org/display/DW/ONAP+on+Kubernetes#ONAPonKubernetes-HardwareRequirements>, <http://onap.readthedocs.io/en/latest/guides/onap-developer/settingup/fullonap.html#requirements>, including one interesting spreadsheet here: <https://lists.onap.org/pipermail/onap-discuss/2017-December/006702.html>. I'm wondering why ONAP requires so much memory? Is there any long term plan to optimize and reduce ONAP's hardware resource requirements? Thanks, Ing-Wher Chen _______________________________________________ onap-discuss mailing list onap-discuss@lists.onap.org https://lists.onap.org/mailman/listinfo/onap-discuss This message and the information contained herein is proprietary and confidential and subject to the Amdocs policy statement, you may review at https://www.amdocs.com/about/email-disclaimer <https://www.amdocs.com/about/email-disclaimer>
_______________________________________________ onap-discuss mailing list onap-discuss@lists.onap.org https://lists.onap.org/mailman/listinfo/onap-discuss
