Kern Sibbald wrote (2006/12/05): > > I used the staging in Amanda too, but since I switched to Bacula (with > > mid-step over afbackup), I do not want it anymore. > You cited why you didn't like Amanda, but I would be interested to know why > you didn't like afbackup, and for the other users who have answered this
I'm very dependent on the feature, which detects holes in files, so that I do not backup or even worse restore one empty 4 TB file on a disk with just 4 GB. Afbackup didn't do that and there was not any indication over the long time (about two years I think), that there is any active development, which was the same problem as with Amanda. As in Amanda case, Afbackup does not have real database for stored files too, there are just listings in text files, and file positions were very inaccurate (up to several tape files before the real position), which was frustrating during restore process - will it be found, or not? Another problem I remember was with spooling, which I take for high end tapes as a must - afbackup had just mmap(), which is limited to 2 GB on FreeBSD and it is memory/swap backed and I could not imagine any scenario, that I can constantly feed LTO3 drive for a sufficient time using this small spool space. The inherent problem was with parallel backups, where files from various jobs are interleaved. It is good for backup time, but a killer for restore time, I tried it just once and please no more :o) >From a long time view, I take as a big advance of Bacula, that there is some view from above and there is visible software architecture, so that I see big potential in further development of Bacula. In the Amanda and afbackup, I did not notice such any thing. > thread, I would be very interested to know why you don't like Areika. Having > this information always helps ... Areika is a new tool for me - I did not know about this software up to now, so I'm interested about it too. I seen just German pages. > Interesting, but not too surprising since LTO-3 is really fast and maybe CPUs > are not quite up to it yet. It seems to me that the problem is with even very small delays (excessive locking can be probably a problem too), where tape drive expects data write requests as fast as possible after ack from previous writes, not counting that their speed would accomodate just limited data rate. In the case, where there are any delays, it seems to me that tape artifically slows down incoming data stream just for prevention, so that initiator can with the same delay write another data block. However it is not commited by any official source, it is just based on my experience and my reasoning. > With technology like LTO-3, perhaps it is time > to consider giving the user the ability to turn > off the CRC32 generation and comparison. It would be very good thing. I tried to patch Bacula personally, but then I was afraid to use it in my real environment, until it is tested and until I'm sure I can restore the data too :o) The patches were not finished and maybe they are lost. I decided that I'm not sufficiently ready to do such a low level thing. I have read about LTO3, that "the tapes contain a strong error correction algorithm that makes data recovery possible when lost data is within one track or up to 32 mm of the tape medium." Another interesting thing is that LTO3 tape drive uses flying tape over the heads (like in hard drives) as much as possible, so cleaning tapes are almost not used in effect. HP says that the drive can run up to 6 000 hours without cleaning tape use (where just small internal brush is sufficient), which was big surprise for me. Unfortunatelly I have got this information after I have bought 5 cleaning tapes :o) Imagine, that one cleaning tape may be exhausted even up to after 90 000 hours = 3750 days >= 10 years... :o) -- Rudolf Cejka <cejkar at fit.vutbr.cz> http://www.fit.vutbr.cz/~cejkar Brno University of Technology, Faculty of Information Technology Bozetechova 2, 612 66 Brno, Czech Republic ------------------------------------------------------------------------- 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.php&p=sourceforge&CID=DEVDEV _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
