Re: [Bacula-users] disk spooling and despooling: concurrency
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Alan Brown wrote: On Sat, 24 Mar 2007, Marc Cuypers wrote: I'm asking this because backups without spooling are faster than those with spooling. Is this normal? Yes - for a single job If you have multiple concurrent jobs running the individual backups will run as slow as a single backup, but because they will interleave you are able to do 2 backups in the same amount of time as one backup. And something I have not seen said yet in this thread: NOT using spooling for concurrent backups will cause restores on any one of those jobs to take exponentially longer due to interleaved data. A penny saved is a dollar earned. :) - -- _ _ _ _ ___ _ _ _ |Y#| | | |\/| | \ |\ | | |Ryan Novosielski - Systems Programmer III |$| |__| | | |__/ | \| _| |[EMAIL PROTECTED] - 973/972.0922 (2-0922) \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.5 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGERwgmb+gadEcsb4RAiNZAJ9adsPlLZwSNBolko5zSs7RKIDD0QCgxwx2 8Gtc22O8iyNEJ6Zamepleio= =CzrC -END PGP SIGNATURE- - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] disk spooling and despooling: concurrency
Hi, On 3/25/2007 5:39 PM, Marc Cuypers wrote: Hi Arno, Thanks for answering. Arno Lehmann wrote: Hi, On 3/24/2007 9:16 PM, Marc Cuypers wrote: Hi, I'm still using bacula 1.38.5. To prevent shoe shining (writing, rewinding, writing again) with a tape drive bacula uses spooling. ... The spooling to the file and the writing to tape is not concurrent. Is there a reason for this? Yes, but what it boils down to is that noone implemented a buffering scheme like using a circular buffer and filling and emptying it simultaneously. Note that many hard disk systems have already difficulties delivering data fast enough to allow streaming on todays tape drives. That would get worse if, due to concurrent writes and reads to one disk, access times increased. Has this changed in later versions? No, not as far as I know. Are there any changes to be expected in this? I doubt it. I'm asking this because backups without spooling are faster than those with spooling. Is this normal? It can be normal, and in most cases is. Overall throughput will probably increase when you run multiple concurrent jobs. When running concurrent jobs, each job will have its own spool file in the spooling directory? And normally when a spool file fills for one job, writing to tape can begin, while the other jobs keep spooling? Alan answered this already, so I just add my comment because you addressed me :-) Yes and yes. Arno -- Best regards, marc Questionable day. Ask somebody something. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- IT-Service Lehmann[EMAIL PROTECTED] Arno Lehmann http://www.its-lehmann.de - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] disk spooling and despooling: concurrency
On Sat, 24 Mar 2007, Marc Cuypers wrote: I'm asking this because backups without spooling are faster than those with spooling. Is this normal? Yes - for a single job If you have multiple concurrent jobs running the individual backups will run as slow as a single backup, but because they will interleave you are able to do 2 backups in the same amount of time as one backup. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] disk spooling and despooling: concurrency
Hi Arno, Thanks for answering. Arno Lehmann wrote: Hi, On 3/24/2007 9:16 PM, Marc Cuypers wrote: Hi, I'm still using bacula 1.38.5. To prevent shoe shining (writing, rewinding, writing again) with a tape drive bacula uses spooling. ... The spooling to the file and the writing to tape is not concurrent. Is there a reason for this? Yes, but what it boils down to is that noone implemented a buffering scheme like using a circular buffer and filling and emptying it simultaneously. Note that many hard disk systems have already difficulties delivering data fast enough to allow streaming on todays tape drives. That would get worse if, due to concurrent writes and reads to one disk, access times increased. Has this changed in later versions? No, not as far as I know. Are there any changes to be expected in this? I doubt it. I'm asking this because backups without spooling are faster than those with spooling. Is this normal? It can be normal, and in most cases is. Overall throughput will probably increase when you run multiple concurrent jobs. When running concurrent jobs, each job will have its own spool file in the spooling directory? And normally when a spool file fills for one job, writing to tape can begin, while the other jobs keep spooling? -- Best regards, marc Questionable day. Ask somebody something. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] disk spooling and despooling: concurrency
Hi, I'm still using bacula 1.38.5. To prevent shoe shining (writing, rewinding, writing again) with a tape drive bacula uses spooling. When using spooling, the data is first spooled to local disk file, then it is written to tape (despooling). When there is more data to backup then the maximum spool file size, this process is repeated as many times as needed. The spooling to the file and the writing to tape is not concurrent. Is there a reason for this? Has this changed in later versions? Are there any changes to be expected in this? I'm asking this because backups without spooling are faster than those with spooling. Is this normal? -- Best regards, marc - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] disk spooling and despooling: concurrency
Hi, On 3/24/2007 9:16 PM, Marc Cuypers wrote: Hi, I'm still using bacula 1.38.5. To prevent shoe shining (writing, rewinding, writing again) with a tape drive bacula uses spooling. ... The spooling to the file and the writing to tape is not concurrent. Is there a reason for this? Yes, but what it boils down to is that noone implemented a buffering scheme like using a circular buffer and filling and emptying it simultaneously. Note that many hard disk systems have already difficulties delivering data fast enough to allow streaming on todays tape drives. That would get worse if, due to concurrent writes and reads to one disk, access times increased. Has this changed in later versions? No, not as far as I know. Are there any changes to be expected in this? I doubt it. I'm asking this because backups without spooling are faster than those with spooling. Is this normal? It can be normal, and in most cases is. Overall throughput will probably increase when you run multiple concurrent jobs. Arno -- IT-Service Lehmann[EMAIL PROTECTED] Arno Lehmann http://www.its-lehmann.de - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users