Re: Provide list of files changed to avoid the crawl
Thanks Jean. On Fri, Jan 24, 2014 at 9:58 PM, Jean-Louis Martineau wrote: > Varun, > > You could add a pre-dle-estimate/pre-dle-backup script that put the list > of files in a file and configure an 'exclude list' for amgtar to use that > file. > > If you backup only files (no directory) then amanda will never notice when > files are removed, so a restore will also restore removed files. > > Jean-Louis > > > On 01/24/2014 09:09 AM, Varun Shastry wrote: > >> Hi experts, >> >> I came to know that amanda client/agent crawls the file system under the >> given path to get the list of files that have changed since the previous >> backup. But in my use case because of the file system size (in several TBs >> using any of the clustered file system) I can't afford crawl every time. >> But there is a utility I use, which gives me list of files changed for any >> given two points in time. My main question here is that, >> >> Is there any method I could use to provide the list of files changed to >> the amanda client so that I can avoid amanda from crawling? If its there >> can I have some pointers to go through it? >> >> Thanks >> Varun Shastry >> >> PS: I agree, for full fledged backup crawling suggested, but with the >> above method I can at least avoid crawling for differential/incremental >> backups. >> > >
Re: Provide list of files changed to avoid the crawl
Varun, You could add a pre-dle-estimate/pre-dle-backup script that put the list of files in a file and configure an 'exclude list' for amgtar to use that file. If you backup only files (no directory) then amanda will never notice when files are removed, so a restore will also restore removed files. Jean-Louis On 01/24/2014 09:09 AM, Varun Shastry wrote: Hi experts, I came to know that amanda client/agent crawls the file system under the given path to get the list of files that have changed since the previous backup. But in my use case because of the file system size (in several TBs using any of the clustered file system) I can't afford crawl every time. But there is a utility I use, which gives me list of files changed for any given two points in time. My main question here is that, Is there any method I could use to provide the list of files changed to the amanda client so that I can avoid amanda from crawling? If its there can I have some pointers to go through it? Thanks Varun Shastry PS: I agree, for full fledged backup crawling suggested, but with the above method I can at least avoid crawling for differential/incremental backups.
Provide list of files changed to avoid the crawl
Hi experts, I came to know that amanda client/agent crawls the file system under the given path to get the list of files that have changed since the previous backup. But in my use case because of the file system size (in several TBs using any of the clustered file system) I can't afford crawl every time. But there is a utility I use, which gives me list of files changed for any given two points in time. My main question here is that, Is there any method I could use to provide the list of files changed to the amanda client so that I can avoid amanda from crawling? If its there can I have some pointers to go through it? Thanks Varun Shastry PS: I agree, for full fledged backup crawling suggested, but with the above method I can at least avoid crawling for differential/incremental backups.
