Hello again all, I've tried looking in the manual for help, but I've found that while the manual is fairly comprehensive, it's not remarkably accessible; that is, while I'm sure my answers are probably in there somewhere, I'm not at all sure that I could find them, so I come to you for help. On this page, http://www.bacula.org/rel-manual/Automated_Disk_Backup.html, the manual describes one way to do disk backups. It seems very straightforward and easy to understand. For simplicity, it only backs up a single client. It does not, however, describe how to add more clients to make a practical system. I would like, if possible, some clarification as to the effect of the "Maximum Volume Jobs" directive. The documentation gives the following:
"http://www.bacula.org/rel-manual/FileSet_Resource.html#PoolResource Maximum Volume Jobs = <positive-integer> This directive specifies the maximum number of Jobs that can be written to the Volume. If you specify zero (the default), there is no limit. Otherwise, when the number of Jobs backed up to the Volume equals positive-integer the Volume will be marked Used. When the Volume is marked Used it can no longer be used for appending Jobs, much like the Full status but it can be recycled if recycling is enabled, and thus used again." That seems to make sense, but then I remember that each client has to have its own Job stanza in bacula-dir.conf. That makes me wonder tonight's backup of each of my clients (call it 4) will take up one job on the volume. This is obviously not the desired behavior, so I look for information on jobs: "http://www.bacula.org/rel-manual/Configuring_Director.html#SECTION001430000000000000000 The Job resource defines a Job (Backup, Restore, ...) that Bacula must perform. Each Job resource definition contains the name of a Client and a FileSet to backup, the Schedule for the Job, where the data are to be stored, and what media Pool can be used. In effect, each Job resource must specify What, Where, How, and When or FileSet, Storage, Backup/Restore/Level, and Schedule respectively. Note, the FileSet must be specified for a restore job for historical reasons, but it is no longer used. Only a single type (Backup, Restore, ...) can be specified for any job. If you want to backup multiple FileSets on the same Client or multiple Clients, you must define a Job for each one. Note, you define only a single Job to do the Full, Differential, and Incremental backups since the different backup levels are tied together by a unique Job name. Normally, you will have only one Job per Client, but if a client has a really huge number of files (more than several million), you might want to split it into to Jobs each with a different FileSet covering only part of the total files. Pool = <pool-resource-name> The Pool directive defines the pool of Volumes where your data can be backed up. Many Bacula installations will use only the Default pool. However, if you want to specify a different set of Volumes for different Clients or different Jobs, you will probably want to use Pools. For additional details, see the Pool Resource section of this chapter. This directive is required." This says nothing about how they interact with volumes, but to me suggests that each client's backup will be counted as a job. My question then becomes: when I want to add more clients to the example on the Automated Disk Backup page, how do I need to modify the configuration files given? Do I just need to add new Client and Job stanzas to each, or do I need to tweak the values in the Pool stanzas to reflect the number of clients (a simple multiplication)? I have asked this question at least twice on this list and got a vague answer the first time and blown off the second. If I could find my answers in the manual I would not be asking the list for help. Thank you, Kyle Marsh > > > > [Scott] was asking how one should configure Bacula to make the best > > > > use of his space. I looked at the example in the manual, here: > > > > http://www.bacula.org/rel-manual/Automated_Disk_Backup.html, but it > > > > seems that is simply an example -- as far as I can tell it only works > > > > for a single client and is not easily scalable, which in impractical. > > > > Tell me if I'm wrong. > > > > > > I'm not exactly sure where you get the above ideas, because I don't > > > believe they are correct -- as you requested, I'm telling you :-) > > > > > > Yes, it is an example, but an example of a *real* installation that has > > > been running without intervention since 19 July 2004 (the exception is > > >that I have > > > upgraded Bacula from time to time, and I had to add a harddisk after the > > > first year as their data volume temporarily doubled). > > > > > > Thought the example is only for a single client, I'm not sure why you say > it > > > works with one. To the best of my knowledge there is no such limitation. > > > Depending on the backup volume you may need to use different Volumes sizes > > > and a different number of Volumes. The scheme has *absolutely* nothing > > > making it inherently single client. > > > > So what does the "Maximum Volume Jobs" directive actually mean, then? > > If I have a single pool for incremental backups which specifies > > Maximum Volume Jobs = 6, > > Maximum Volumes = 6, and Volume Retention = 14 days, when I back up 4 > > clients each night, what happens? Does it say "The 4 incremental > > backups coming in tonight are from different clients so they're all > > part of the same job and can go on a single volume as one job" or does > > it say "The four backups coming in tonight have different Job stanzas > > so they must be different jobs so they'll take up 4 of the 6 job slots > > in the current volume that was supposed to last a week," or does it do > > something else? I've asked this question before, and the answer I got > > suggested I do something else (see below) suggesting that something > > like option 2 was the case. Could you clear this up? > > I think the above terms are quite well spelled out in the manual. If you are > still confused, I recommend you ask on the bacula-users list. > > As I said there are many ways to do things. I see absolutely nothing wrong > with Arno's suggestions. > > I'll read the documentation of that chapter over again tomorrow to be sure it > is clear how you might want to adjust things if you added another client, but > I thought I had explained the theory behind it so that you could modify the > parameters to correspond to what you want. > > > > > > Thanks and sorry for the trouble, > > > > ~Kyle Marsh > > > > > > > > Concerning scalability: As with anything, if you double the size, it is > > > probably no problem, but if you multiply by 10 there may be additional > > > factors that come it -- for example, I am using SQLite2 (a very old > version) > > > and anything more serious (in terms of volume) I would use MySQL or > > > PostgeSQL. As such this scheme has plenty of room to scale. Up to about > 50 > > > machines, I could imagine just adding more disk (and the db change I > > > mentioned). Beyond that, one would likely need to make other adjustments > > > that would be made no matter what scheme one chooses. > > > > > > There are many different ways to handle disk backups (with/without pools, > > > different schedules of Full/differential/incremental, ...). The *working* > > > example I gave in the manual is only one way to do it, but it does work. > > > > > > I hope I have dispelled any ideas that there is any fundamental > limitations to > > > the scheme -- though after 3 years of continuous running, I could probably > > > find a few tweaks to make it better ... > > > > > > Oh yes, and once someone blew away his whole home directory while I was on > > > vacation. I got a call on my mobile, and in ten minutes they had the > > > directory back. > > > > > > > > > Regards, > > > > > > Kern > > > > > > > > > > > > > I got advice from Arno Lehman suggesting that I estimate the amount of > > > > space needed per pool and set up the pool to handle that amount of > > > > data using limits on volume size and number of volumes. My reply was > > > > sent out with the wrong address so the list moderator ate it and I > > > > haven't resent it to ask my clarifying questions. > > > > > > > I too wonder what would be a reasonable volume size given various pool > > > > sizes (I'm thinking about one pool for each level of backups with its > > > > own retention times running off the basic monthly cycle that comes > > > > preconfigured). > > > > > > > Thank you, > > > > > > > ~Kyle Marsh > > > > > > ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
