[Bacula-users] Database performance issues
Hi All, We have a setup with bacula-dir/MySQL and bacula-sd (2 servers) and a SAN connected via fiber. We are running bacula with about 250 jobs a day. The table File is about 500M records with a size of about 100G. We use batch insert. All tables are MyISAM. Currently we are experiencing performance problems with the database. For all jobs we do a disk-to-disk backup. The next day we use a copyjob to put the data on tape. During daytime we run the copy jobs and during the evening/night we run the disk-to-disk jobs. There is some overlap. When the the disk-to-disk jobs start, we see number of queries on the database drop to virtually 0. There are a few options to solve this - use innodb for the tables in MySQL - migrate to PostgreSQL Can anyone advise me what to do, to increase the performance of our platform? Kind regards, Enrico van Goor -- EditLive Enterprise is the world's most technically advanced content authoring tool. Experience the power of Track Changes, Inline Image Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Multiple Next Pools
Phil, On Wed, 2010-07-21 at 12:05 -0400, Phil Stracchino wrote: > On 07/21/10 09:26, Enrico van Goor wrote: > > Hi all, > > > > We make backups of about 200 server at the moment, to disk. Twice a week > > a Full backup and 5 times a week an incremental. Every day is a > > different pool, so the tape drives won't interfere with the backup to > > disk and vice versa. We write the data to tape using a copy job. > > > > The tapes have a retention of 30 days. Now we want to archive some data > > forever. Lets say 100 years. The thought was, since we already have the > > data on disk I wanted to use a second copy job to copy the data to a > > separate pool, where we use a Migrate job to move the data to a > > different tape set. > > > > So what I basically want is that data is copied on a daily basis to > > pool_a and once a month to pool_a and pool_b. > > > > I can't figure out how to configure this. I don't know if it is even > > possible. Can anyone help me with this? > > Can you not simply create a copy job with destination of pool_a that > runs once daily, and another with destination of pool_b on a > once-a-month schedule? Well, pool_a are 2 tapedrives. In the near future we have a very tight schedule for both the tapedrives. They probably will be taping about 23 hours a day (1 hour maintenance window). So I don't think it is a good idea to get the data from tape again. And you have the risk that the tape isn't in the robot anymore. Kind regards, Enrico van Goor System & Network Engineer -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] Multiple Next Pools
Hi all, We make backups of about 200 server at the moment, to disk. Twice a week a Full backup and 5 times a week an incremental. Every day is a different pool, so the tape drives won't interfere with the backup to disk and vice versa. We write the data to tape using a copy job. The tapes have a retention of 30 days. Now we want to archive some data forever. Lets say 100 years. The thought was, since we already have the data on disk I wanted to use a second copy job to copy the data to a separate pool, where we use a Migrate job to move the data to a different tape set. So what I basically want is that data is copied on a daily basis to pool_a and once a month to pool_a and pool_b. I can't figure out how to configure this. I don't know if it is even possible. Can anyone help me with this? Kind regards, Enrico van Goor System & Network Engineer. -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] waiting on Storage
Hi All, I'm having problems with the scheduling of jobs. We have a setup with a tape robot with 2 drives and currently about 200 servers to backup. We make a backup of a server to disk. The next day it is copied to Tape with a copy job. To make optimal use of the tapedrives, we created 14 disk pools (2 for every day) and 28 schedules (4 for every day: 2 disk pools and 2 per disk pool to devide the full backups better). The problem occurs when� there is a Full backup of a servers which has a lot of data. The job runs for more then an hour. The jobs in the same schedule are run concurrent (set to 12). But when the next schedule starts, those jobs wait on the large job to be finished with the message waiting on Storage. When issueing the command 'echo "status stor" | bconsole' the next thing caught our attention: Jobs waiting to reserve a drive: �� 3608 JobId=166029 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166033 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166035 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166037 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166041 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166043 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166044 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166045 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166053 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166056 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). �� 3608 JobId=166057 wants Pool="Disk_wed_2-Pool" but have Pool="" nreserve=0 on drive "diskbackup_wed_2" (/bacula/wednesday/diskbackup_wed_2). The jobs have no pool. Does anyone know how to solve this? � Kind regards, Enrico van Goor System & Network Engineer -- ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
