On 12/02/2014 1:40 AM, Peter Palfrader wrote:
On Wed, 29 Jan 2014, James Bromberger wrote:Peter, does this sound sensible - the logic for restores, and the setting of content-disposition header so clients can download direct from the AWS S3 Bucket?I can totally see that being useful as a backup. I'm still less than convinced that offline files are suitable for something that people have in their sources.list or browse interactively.
The objective is to keep the last X days with on live storage, and the prior history 'near line' (3-5 hours). At the moment I have X at 365 days (after 7 January when I finished importing historical files). Perhaps we try and ensure that 90/95/99% of files accessed are on live storage by adjusting this time-to-archive value; but in doing so, the cost is minimised. If X needs to be 4 years then that's fine.
Sure. I've got in to it fine (thanks for access) - I see the farm-journal has recent uploads (2 days worth) so i can script around that to ensure we copy across.Peter, is S3 receiving a copy of all new ingested files into snapshot? If not, who should do this?I suspect the easiest approach might be a new cronjob as snapshot user on sibelius that does that?
I thought that we ended on the crypt-in-main was precedent for the archive containing all files.Have you figured out how to handle non-US yet?
Also, may I get a read-only access to the live postgres DB to be able to monitor files being ingested and to check the real filenames that should be set (access from localhost on sibelius is fine)?See other email.
Thanks again. James