This functionality has been discussed on the -devel list recently there are
plans to implement it.  Unfortunately, there is no specific timetable on the
subject as of yet.  In short, "It's coming...eventually." :-)

Thanks,
Chris  

> -----Original Message-----
> From: [EMAIL PROTECTED] 
> [mailto:[EMAIL PROTECTED] On Behalf 
> Of Alexander Kuehn
> Sent: Tuesday, 30 August, 2005 08:03
> To: bacula-users@lists.sourceforge.net
> Subject: [Bacula-users] parallel spooling/despooling
> 
> Hi,
> 
> I have a [EMAIL PROTECTED] FreeBSD machine with a DDS3 Streamer 
> attached to it. The streamer does about 1MB/s, when I use 
> Bacula I can monitor the machine and streamer usage using 
> "systat -vm", when I do that I see that nothing is running at 
> full steam, the streamer does about 0.8 MB/s, the CPU is at 
> 80% usage, the harddisk is 5-20% usage. Of course this varies 
> over time, but these values are pretty typical. I was 
> wondering if I could optimize the usage to minimize the 
> backup times. I was very happy to see that Bacula supports 
> spooling at first, but I was kinda set back when I saw that 
> spooling and despooling is done sequentially not in parallel. 
> I had already set up a mfs based /tmp (128MB) to have Bacula 
> spool to RAM (which the machine has plenty of), and CPU usage 
> was indeed 100% as I had expected. Of course the CPU was 
> basically unutilized when it came to unspooling the file.
> I wonder if it would be difficult to modify Bacula to use two 
> spool files instead of one, so that:
> fd starts collecting files and creating the normal spool file,
> as soon as it is ready and signals sd to unspool the file, it 
> does not wait until sd has finished but continues to collect 
> files but stores them in a second file, right next to the 
> normal spool file,
> as soon as sd has finished unspooling and fd spooling to the 
> second file, the second file is renamed to the normal spool 
> file and fd starts to collect for a new second file...
> I think this way the backup time could be significantly 
> reduced in cases such as this one.
> Kind regards,
> Alex.
> 
> 
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development 
> Lifecycle Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * 
> Testing & QA
> Security * Process Improvement & Measurement * 
> http://www.sqe.com/bsce5sf
> _______________________________________________
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users



-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Reply via email to