| 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 <2016> 2017 2018 2019 2020 2021 2022 2023 2024 2025 | Index | 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 <2016> 2017 2018 2019 2020 2021 2022 2023 2024 2025 |
| <== Date ==> | <== Thread ==> |
|---|
| Subject: | RE: Archive Appliance Backends |
| From: | Mazanec Tomáš <[email protected]> |
| To: | "Kasemir, Kay" <[email protected]>, William Cork <[email protected]>, "[email protected]" <[email protected]> |
| Date: | Wed, 21 Sep 2016 14:37:29 +0000 |
|
Hi
Footnote on Cassandra Archiver from a happy user: Cassandra Archiver has been superseded by "Cassandra PV Archiver" http://oss.aquenos.com/cassandra-pv-archiver/ Cassandra PV Archiver does not inherit from RDB Archiver (as its predecessor did) but it was newly designed from scratch. -- it is pluggable in terms of that user can write custom control system support which implements front-end and back-end interfaces. Release v 3.0.x reads Channel Access PVs and writes to Apache Cassandra cluster. -- built-in retrieval interface offers JSON (and there is CSS plugin counterpart) -- It implements retention and compression of archived samples -- it is designed to be deployed as a cluster of engines which cooperate on-line. It is horizontally scalable. Cheers, Tomas Od: [email protected] [[email protected]] za uživatele Kasemir, Kay [[email protected]]
Odesláno: 21. září 2016 12:57 Komu: William Cork; [email protected] Předmět: Re: Archive Appliance Backends Hi:
>This is my first EPICS tech-talk post.
Welcome
>Has anyone created an alternative backend to PB files?
>I'm looking into utilising
a time series database instead
Good question.
I am not aware of that being available for the Archive Appliance.
The archive appliance code started out using the code for the archive engine that's in the CS-Studio repository, see
That one originally uses a relational database (MySQL, Oracle or postgres) for data storage, which is very reliable but not fast.
The code that actually writes to the RDB is in plugs that can be replaced with another type of storage.
There is, for example, an implementation that uses Apache Cassandra, see
The archive appliance has by now deviated quite a lot from the original archive engine that's in the CS-Studio repository because it has a much nicer web
interface and some amount of self-maintenance.
I don't know if this means that the underlying storage is still somewhat "pluggable" so that you could replace the custom
collection of protobuf files with for example Cassandra.
Thanks,
Kay
|