Re: [Bacula-users] PostgreSQL: WAL and log management
Thank you. -- George ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] PostgreSQL: WAL and log management
On 11/26/2018 11:40 AM, George via Bacula-users wrote: > Just one more question: when automating the backup of > all postgresql databases, should I also backup > template1? I have found that template0 cannot be > backed up due to how postgresql works but couldn't > find similar info about template1. I don't bother with its built-in databases: IME reinstalling postgres (which restores them) is as fast or faster than restoring from backup. If you're after high availability-level of backup, set up another server and run streaming replication to it. -- Dimitri Maziuk Programmer/sysadmin BioMagResBank, UW-Madison -- http://www.bmrb.wisc.edu signature.asc Description: OpenPGP digital signature ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] PostgreSQL: WAL and log management
Thanks a lot for your help Dmitri. I decided to start from scratch for the sake of exploring how this would work without repeating many times the same things. So I: 1. Made a backup of /var/lib/pgsql 2. Stopped service and uninstalled postgresql 3. Removed /var/lib/pgsql 4. Reinstalled postgresql 5. Applied the conf changes discussed here 6. Imported the SQL data for bacula Right now data/pg_wal is 81M (5x16M) files in it. And data/base is 126M of which 105M seem to belong to bacula database. Just one more question: when automating the backup of all postgresql databases, should I also backup template1? I have found that template0 cannot be backed up due to how postgresql works but couldn't find similar info about template1. -- George ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] PostgreSQL: WAL and log management
On Sun, 25 Nov 2018 21:16:03 +0200 George via Bacula-users wrote: > The WAL files remain though. Should I delete them > manually? Postgres will fail to start if they're actually used. I think there are some functions to figure out which ones are used but I suppose you could just move them to e.g. /var/tmp and tehn move them back one by one starting from newest until it starts. > I suppose it may be due to the > multiple imports and table deletions. But it still > looks too much considering that the total .sql file is > less than 70MB. Not really, I know that for e.g. deletions, a WAL stores the entire old row, and the entire new row. You could run a bunch of deletes and inserts and make it grow big, that's how yours could get that way. The issue though is that posgres is largely self-tuning, and normally that would gradually go down to max_wal_size. -- Dmitri Maziuk ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] bacula copy job fails when tape is full and needs to change
Thanks Kern! I will dig deeper in documentation. First I try setup heartbeat in SD's config! On Mon, Nov 26, 2018 at 11:33 AM Kern Sibbald wrote: > Oh there are at least five different places to setup Heart Beat Interval > (Dir, SD, and FD). Unfortunately my memory is not good enough to remember > them all. Please ask others or see the documentation ... > > The easiest way is to get on a current version -- e.g. 9.2.2 where it is > done by defaut. > > Best regards, > Kern > > On 11/26/18 11:13 AM, Andras Horvai wrote: > > Hello Kern, > > yes you are right I am using bacula 7.0.5 shipped with Ubuntu 16.04. > Where should I setup heartbeat interval? On SD's or FD's config? Or both? > > Thanks for your help! > > Andras > > On Mon, Nov 26, 2018 at 10:56 AM Kern Sibbald wrote: > >> Hello, >> >> If I remember right you are running on a *very* old Bacula, and the >> problem seems to be that the backup takes more than 2 hours. One of your >> comm lines (SD <-> FD) times out. I mention your old version because newer >> Bacula's automatically fix this problem by turning on Heart Beat Interval = >> 300, which is very likely to resolve your problem. >> >> Best regards, >> Kern >> >> On 11/26/18 10:34 AM, Andras Horvai wrote: >> >> Hi Tilman, >> >> thank you for your answer! But unfortunately the firewall cannot be a >> problem here :) >> The problem happens only with Copy Jobs. The SD and the FD is on the same >> device. There is no firewall on the machine. >> So what I am doing is the following: >> >> during weekend I do full backup with the backup server to file storage on >> the backup server. Then starting from Monday I am doing Copy Job from the >> backup >> server to a Tape device connected to the backup server. This works pretty >> well till Tape does not get full. When Tape gets full bacula asks for >> another tape. >> We replace the tape, so job would continue (as expected) but then at the >> end we got the job error... So I am puzzled what is wrong. >> >> Please feel free to share your ideas... >> >> Thanks, >> >> Andras >> >> On Sun, Nov 25, 2018 at 10:28 PM Tilman Schmidt >> wrote: >> >>> Hi Andras, >>> >>> is there a firewall between the client and the SD? >>> The message >>> >>> > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network >>> error reading from FD. ERR=Connection reset by peer >>> >>> looks suspiciously like a firewall killing the FD - SD connection >>> because it sees it as idle. >>> >>> HTH >>> Tilman >>> >>> Am 22.11.2018 um 16:04 schrieb Andras Horvai: >>> > Dear list, >>> > >>> > I have to following problem: >>> > We use copy jobs to copy weekly full backups to WORM tape but when a >>> tape >>> > gets filled and needs to change the copy job failed. Bacula says >>> > intervention is >>> > needed so we put a new tape in the tape drive. What can be the problem? >>> > >>> > Copy job report: >>> > 20-Nov 12:25 backup-sd JobId 4838: End of Volume at file 0 on device >>> > "FileStorage" (/backup), Volume "FILEW-0542" >>> > 20-Nov 12:25 backup-sd JobId 4838: End of all volumes. >>> > 20-Nov 12:25 backup-sd JobId 4838: Elapsed time=02:45:29, Transfer >>> > rate=42.14 M Bytes/second >>> > 20-Nov 12:25 backup1 JobId 4838: Error: Bacula backup1 7.0.5 (28Jul14): >>> > Build OS: x86_64-pc-linux-gnu ubuntu 16.04 >>> > Prev Backup JobId: 4837 >>> > Prev Backup Job:db1-job.2018-11-19_23.09.19_03 >>> > New Backup JobId: 4845 >>> > Current JobId: 4838 >>> > Current Job:Archive.2018-11-20_07.59.53_05 >>> > Backup Level: Full >>> > Client: None >>> > FileSet:"None" 2017-06-19 09:00:00 >>> > Read Pool: "ServersWeeklyFullFile" (From Job resource) >>> > Read Storage: "File" (From Pool resource) >>> > Write Pool: "TapeArchive" (From Pool's NextPool resource) >>> > Write Storage: "LTO-4" 20-Nov 09:39 backup1 JobId 4845: >>> Using >>> > Device "LTO-4" to write. >>> > 20-Nov 12:01 backup-sd JobId 4845: End of Volume "WORMW-1242" at >>> > 386:27137 on device "LTO-4" (/dev/nst0). Write of 64512 bytes got -1. >>> > 20-Nov 12:01 backup-sd JobId 4845: Re-read of last block succeeded. >>> > 20-Nov 12:01 backup-sd JobId 4845: End of medium on Volume "WORMW-1242" >>> > Bytes=764,853,046,272 Blocks=11,855,980 at 20-Nov-2018 12:01. >>> > 20-Nov 12:01 backup1 JobId 4845: Created new Volume="WORMW-1243", >>> > Pool="TapeArchive", MediaType="LTO-4" in catalog. >>> > 20-Nov 12:01 backup-sd JobId 4845: Please mount append Volume >>> > "WORMW-1243" or label a new one for: >>> > Job: db1-job.2018-11-20_07.59.54_12 >>> > Storage: "LTO-4" (/dev/nst0) >>> > Pool: TapeArchive >>> > Media type: LTO-4 >>> > 20-Nov 12:15 backup-sd JobId 4845: Error: The Volume=WORMW-1243 on >>> > device="LTO-4" (/dev/nst0) appears to be unlabeled. >>> > 20-Nov 12:15 backup-sd JobId 4845: Labeled new Volume "WORMW-1243" on >>> >
Re: [Bacula-users] bacula copy job fails when tape is full and needs to change
Oh there are at least five different places to setup Heart Beat Interval (Dir, SD, and FD). Unfortunately my memory is not good enough to remember them all. Please ask others or see the documentation ... The easiest way is to get on a current version -- e.g. 9.2.2 where it is done by defaut. Best regards, Kern On 11/26/18 11:13 AM, Andras Horvai wrote: Hello Kern, yes you are right I am using bacula 7.0.5 shipped with Ubuntu 16.04. Where should I setup heartbeat interval? On SD's or FD's config? Or both? Thanks for your help! Andras On Mon, Nov 26, 2018 at 10:56 AM Kern Sibbaldwrote: Hello, If I remember right you are running on a *very* old Bacula, and the problem seems to be that the backup takes more than 2 hours. One of your comm lines (SD <-> FD) times out. I mention your old version because newer Bacula's automatically fix this problem by turning on Heart Beat Interval = 300, which is very likely to resolve your problem. Best regards, Kern On 11/26/18 10:34 AM, Andras Horvai wrote: Hi Tilman, thank you for your answer! But unfortunately the firewall cannot be a problem here :) The problem happens only with Copy Jobs. The SD and the FD is on the same device. There is no firewall on the machine. So what I am doing is the following: during weekend I do full backup with the backup server to file storage on the backup server. Then starting from Monday I am doing Copy Job from the backup server to a Tape device connected to the backup server. This works pretty well till Tape does not get full. When Tape gets full bacula asks for another tape. We replace the tape, so job would continue (as expected) but then at the end we got the job error... So I am puzzled what is wrong. Please feel free to share your ideas... Thanks, Andras On Sun, Nov 25, 2018 at 10:28 PM Tilman Schmidt wrote: Hi Andras, is there a firewall between the client and the SD? The message > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network error reading from FD. ERR=Connection reset by peer looks suspiciously like a firewall killing the FD - SD connection because it sees it as idle. HTH Tilman Am 22.11.2018 um 16:04 schrieb Andras Horvai: > Dear list, > > I have to following problem: > We use copy jobs to copy weekly full backups to WORM tape but when a tape > gets filled and needs to change the copy job failed. Bacula says > intervention is > needed so we put a new tape in the tape drive. What can be the problem? > > Copy job report: > 20-Nov 12:25 backup-sd JobId 4838: End of Volume at file 0 on device > "FileStorage" (/backup), Volume "FILEW-0542" > 20-Nov 12:25 backup-sd JobId 4838: End of all volumes. > 20-Nov 12:25 backup-sd JobId 4838: Elapsed time=02:45:29, Transfer > rate=42.14 M Bytes/second > 20-Nov 12:25 backup1 JobId 4838: Error: Bacula backup1 7.0.5 (28Jul14): > Build OS: x86_64-pc-linux-gnu ubuntu 16.04 > Prev Backup JobId: 4837 > Prev Backup Job: db1-job.2018-11-19_23.09.19_03 > New Backup JobId: 4845 > Current
Re: [Bacula-users] bacula copy job fails when tape is full and needs to change
Hello Kern, yes you are right I am using bacula 7.0.5 shipped with Ubuntu 16.04. Where should I setup heartbeat interval? On SD's or FD's config? Or both? Thanks for your help! Andras On Mon, Nov 26, 2018 at 10:56 AM Kern Sibbald wrote: > Hello, > > If I remember right you are running on a *very* old Bacula, and the > problem seems to be that the backup takes more than 2 hours. One of your > comm lines (SD <-> FD) times out. I mention your old version because newer > Bacula's automatically fix this problem by turning on Heart Beat Interval = > 300, which is very likely to resolve your problem. > > Best regards, > Kern > > On 11/26/18 10:34 AM, Andras Horvai wrote: > > Hi Tilman, > > thank you for your answer! But unfortunately the firewall cannot be a > problem here :) > The problem happens only with Copy Jobs. The SD and the FD is on the same > device. There is no firewall on the machine. > So what I am doing is the following: > > during weekend I do full backup with the backup server to file storage on > the backup server. Then starting from Monday I am doing Copy Job from the > backup > server to a Tape device connected to the backup server. This works pretty > well till Tape does not get full. When Tape gets full bacula asks for > another tape. > We replace the tape, so job would continue (as expected) but then at the > end we got the job error... So I am puzzled what is wrong. > > Please feel free to share your ideas... > > Thanks, > > Andras > > On Sun, Nov 25, 2018 at 10:28 PM Tilman Schmidt > wrote: > >> Hi Andras, >> >> is there a firewall between the client and the SD? >> The message >> >> > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network >> error reading from FD. ERR=Connection reset by peer >> >> looks suspiciously like a firewall killing the FD - SD connection >> because it sees it as idle. >> >> HTH >> Tilman >> >> Am 22.11.2018 um 16:04 schrieb Andras Horvai: >> > Dear list, >> > >> > I have to following problem: >> > We use copy jobs to copy weekly full backups to WORM tape but when a >> tape >> > gets filled and needs to change the copy job failed. Bacula says >> > intervention is >> > needed so we put a new tape in the tape drive. What can be the problem? >> > >> > Copy job report: >> > 20-Nov 12:25 backup-sd JobId 4838: End of Volume at file 0 on device >> > "FileStorage" (/backup), Volume "FILEW-0542" >> > 20-Nov 12:25 backup-sd JobId 4838: End of all volumes. >> > 20-Nov 12:25 backup-sd JobId 4838: Elapsed time=02:45:29, Transfer >> > rate=42.14 M Bytes/second >> > 20-Nov 12:25 backup1 JobId 4838: Error: Bacula backup1 7.0.5 (28Jul14): >> > Build OS: x86_64-pc-linux-gnu ubuntu 16.04 >> > Prev Backup JobId: 4837 >> > Prev Backup Job:db1-job.2018-11-19_23.09.19_03 >> > New Backup JobId: 4845 >> > Current JobId: 4838 >> > Current Job:Archive.2018-11-20_07.59.53_05 >> > Backup Level: Full >> > Client: None >> > FileSet:"None" 2017-06-19 09:00:00 >> > Read Pool: "ServersWeeklyFullFile" (From Job resource) >> > Read Storage: "File" (From Pool resource) >> > Write Pool: "TapeArchive" (From Pool's NextPool resource) >> > Write Storage: "LTO-4" 20-Nov 09:39 backup1 JobId 4845: Using >> > Device "LTO-4" to write. >> > 20-Nov 12:01 backup-sd JobId 4845: End of Volume "WORMW-1242" at >> > 386:27137 on device "LTO-4" (/dev/nst0). Write of 64512 bytes got -1. >> > 20-Nov 12:01 backup-sd JobId 4845: Re-read of last block succeeded. >> > 20-Nov 12:01 backup-sd JobId 4845: End of medium on Volume "WORMW-1242" >> > Bytes=764,853,046,272 Blocks=11,855,980 at 20-Nov-2018 12:01. >> > 20-Nov 12:01 backup1 JobId 4845: Created new Volume="WORMW-1243", >> > Pool="TapeArchive", MediaType="LTO-4" in catalog. >> > 20-Nov 12:01 backup-sd JobId 4845: Please mount append Volume >> > "WORMW-1243" or label a new one for: >> > Job: db1-job.2018-11-20_07.59.54_12 >> > Storage: "LTO-4" (/dev/nst0) >> > Pool: TapeArchive >> > Media type: LTO-4 >> > 20-Nov 12:15 backup-sd JobId 4845: Error: The Volume=WORMW-1243 on >> > device="LTO-4" (/dev/nst0) appears to be unlabeled. >> > 20-Nov 12:15 backup-sd JobId 4845: Labeled new Volume "WORMW-1243" on >> > tape device "LTO-4" (/dev/nst0). >> > 20-Nov 12:15 backup-sd JobId 4845: Wrote label to prelabeled Volume >> > "WORMW-1243" on tape device "LTO-4" (/dev/nst0) >> > 20-Nov 12:15 backup-sd JobId 4845: New volume "WORMW-1243" mounted on >> > device "LTO-4" (/dev/nst0) at 20-Nov-2018 12:15. >> > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network >> > error reading from FD. ERR=Connection reset by peer >> > 20-Nov 12:25 backup-sd JobId 4845: Elapsed time=02:31:15, Transfer >> > rate=46.11 M Bytes/second >> > (From Pool's NextPool resource) >> > Catalog:"MyCatalog" (From Client resource) >> > Start time:
Re: [Bacula-users] bacula copy job fails when tape is full and needs to change
Hello, If I remember right you are running on a *very* old Bacula, and the problem seems to be that the backup takes more than 2 hours. One of your comm lines (SD <-> FD) times out. I mention your old version because newer Bacula's automatically fix this problem by turning on Heart Beat Interval = 300, which is very likely to resolve your problem. Best regards, Kern On 11/26/18 10:34 AM, Andras Horvai wrote: Hi Tilman, thank you for your answer! But unfortunately the firewall cannot be a problem here :) The problem happens only with Copy Jobs. The SD and the FD is on the same device. There is no firewall on the machine. So what I am doing is the following: during weekend I do full backup with the backup server to file storage on the backup server. Then starting from Monday I am doing Copy Job from the backup server to a Tape device connected to the backup server. This works pretty well till Tape does not get full. When Tape gets full bacula asks for another tape. We replace the tape, so job would continue (as expected) but then at the end we got the job error... So I am puzzled what is wrong. Please feel free to share your ideas... Thanks, Andras On Sun, Nov 25, 2018 at 10:28 PM Tilman Schmidt wrote: Hi Andras, is there a firewall between the client and the SD? The message > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network error reading from FD. ERR=Connection reset by peer looks suspiciously like a firewall killing the FD - SD connection because it sees it as idle. HTH Tilman Am 22.11.2018 um 16:04 schrieb Andras Horvai: > Dear list, > > I have to following problem: > We use copy jobs to copy weekly full backups to WORM tape but when a tape > gets filled and needs to change the copy job failed. Bacula says > intervention is > needed so we put a new tape in the tape drive. What can be the problem? > > Copy job report: > 20-Nov 12:25 backup-sd JobId 4838: End of Volume at file 0 on device > "FileStorage" (/backup), Volume "FILEW-0542" > 20-Nov 12:25 backup-sd JobId 4838: End of all volumes. > 20-Nov 12:25 backup-sd JobId 4838: Elapsed time=02:45:29, Transfer > rate=42.14 M Bytes/second > 20-Nov 12:25 backup1 JobId 4838: Error: Bacula backup1 7.0.5 (28Jul14): > Build OS: x86_64-pc-linux-gnu ubuntu 16.04 > Prev Backup JobId: 4837 > Prev Backup Job: db1-job.2018-11-19_23.09.19_03 > New Backup JobId: 4845 > Current JobId: 4838 > Current Job: Archive.2018-11-20_07.59.53_05 > Backup Level: Full > Client: None > FileSet: "None" 2017-06-19 09:00:00 > Read Pool: "ServersWeeklyFullFile" (From Job resource) > Read Storage: "File" (From Pool resource) > Write Pool: "TapeArchive" (From Pool's NextPool resource) > Write Storage: "LTO-4" 20-Nov 09:39 backup1 JobId 4845: Using > Device "LTO-4" to write. > 20-Nov 12:01 backup-sd JobId 4845: End of Volume "WORMW-1242" at > 386:27137 on device "LTO-4" (/dev/nst0). Write of 64512 bytes got -1. > 20-Nov 12:01 backup-sd JobId 4845: Re-read of last block succeeded. > 20-Nov 12:01 backup-sd JobId 4845: End of medium on Volume "WORMW-1242" > Bytes=764,853,046,272 Blocks=11,855,980 at 20-Nov-2018 12:01. > 20-Nov 12:01 backup1 JobId 4845: Created new Volume="WORMW-1243", > Pool="TapeArchive", MediaType="LTO-4" in catalog. > 20-Nov 12:01 backup-sd JobId 4845: Please mount append Volume > "WORMW-1243" or label a new one for: > Job: db1-job.2018-11-20_07.59.54_12 > Storage: "LTO-4" (/dev/nst0) > Pool: TapeArchive > Media type: LTO-4 > 20-Nov 12:15 backup-sd JobId 4845: Error: The Volume=WORMW-1243 on > device="LTO-4" (/dev/nst0) appears to be unlabeled. > 20-Nov 12:15 backup-sd JobId 4845: Labeled
Re: [Bacula-users] bacula copy job fails when tape is full and needs to change
Hi Tilman, thank you for your answer! But unfortunately the firewall cannot be a problem here :) The problem happens only with Copy Jobs. The SD and the FD is on the same device. There is no firewall on the machine. So what I am doing is the following: during weekend I do full backup with the backup server to file storage on the backup server. Then starting from Monday I am doing Copy Job from the backup server to a Tape device connected to the backup server. This works pretty well till Tape does not get full. When Tape gets full bacula asks for another tape. We replace the tape, so job would continue (as expected) but then at the end we got the job error... So I am puzzled what is wrong. Please feel free to share your ideas... Thanks, Andras On Sun, Nov 25, 2018 at 10:28 PM Tilman Schmidt wrote: > Hi Andras, > > is there a firewall between the client and the SD? > The message > > > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network > error reading from FD. ERR=Connection reset by peer > > looks suspiciously like a firewall killing the FD - SD connection > because it sees it as idle. > > HTH > Tilman > > Am 22.11.2018 um 16:04 schrieb Andras Horvai: > > Dear list, > > > > I have to following problem: > > We use copy jobs to copy weekly full backups to WORM tape but when a tape > > gets filled and needs to change the copy job failed. Bacula says > > intervention is > > needed so we put a new tape in the tape drive. What can be the problem? > > > > Copy job report: > > 20-Nov 12:25 backup-sd JobId 4838: End of Volume at file 0 on device > > "FileStorage" (/backup), Volume "FILEW-0542" > > 20-Nov 12:25 backup-sd JobId 4838: End of all volumes. > > 20-Nov 12:25 backup-sd JobId 4838: Elapsed time=02:45:29, Transfer > > rate=42.14 M Bytes/second > > 20-Nov 12:25 backup1 JobId 4838: Error: Bacula backup1 7.0.5 (28Jul14): > > Build OS: x86_64-pc-linux-gnu ubuntu 16.04 > > Prev Backup JobId: 4837 > > Prev Backup Job:db1-job.2018-11-19_23.09.19_03 > > New Backup JobId: 4845 > > Current JobId: 4838 > > Current Job:Archive.2018-11-20_07.59.53_05 > > Backup Level: Full > > Client: None > > FileSet:"None" 2017-06-19 09:00:00 > > Read Pool: "ServersWeeklyFullFile" (From Job resource) > > Read Storage: "File" (From Pool resource) > > Write Pool: "TapeArchive" (From Pool's NextPool resource) > > Write Storage: "LTO-4" 20-Nov 09:39 backup1 JobId 4845: Using > > Device "LTO-4" to write. > > 20-Nov 12:01 backup-sd JobId 4845: End of Volume "WORMW-1242" at > > 386:27137 on device "LTO-4" (/dev/nst0). Write of 64512 bytes got -1. > > 20-Nov 12:01 backup-sd JobId 4845: Re-read of last block succeeded. > > 20-Nov 12:01 backup-sd JobId 4845: End of medium on Volume "WORMW-1242" > > Bytes=764,853,046,272 Blocks=11,855,980 at 20-Nov-2018 12:01. > > 20-Nov 12:01 backup1 JobId 4845: Created new Volume="WORMW-1243", > > Pool="TapeArchive", MediaType="LTO-4" in catalog. > > 20-Nov 12:01 backup-sd JobId 4845: Please mount append Volume > > "WORMW-1243" or label a new one for: > > Job: db1-job.2018-11-20_07.59.54_12 > > Storage: "LTO-4" (/dev/nst0) > > Pool: TapeArchive > > Media type: LTO-4 > > 20-Nov 12:15 backup-sd JobId 4845: Error: The Volume=WORMW-1243 on > > device="LTO-4" (/dev/nst0) appears to be unlabeled. > > 20-Nov 12:15 backup-sd JobId 4845: Labeled new Volume "WORMW-1243" on > > tape device "LTO-4" (/dev/nst0). > > 20-Nov 12:15 backup-sd JobId 4845: Wrote label to prelabeled Volume > > "WORMW-1243" on tape device "LTO-4" (/dev/nst0) > > 20-Nov 12:15 backup-sd JobId 4845: New volume "WORMW-1243" mounted on > > device "LTO-4" (/dev/nst0) at 20-Nov-2018 12:15. > > 20-Nov 12:25 backup-sd JobId 4845: Fatal error: append.c:223 Network > > error reading from FD. ERR=Connection reset by peer > > 20-Nov 12:25 backup-sd JobId 4845: Elapsed time=02:31:15, Transfer > > rate=46.11 M Bytes/second > > (From Pool's NextPool resource) > > Catalog:"MyCatalog" (From Client resource) > > Start time: 20-Nov-2018 09:39:31 > > End time: 20-Nov-2018 12:25:04 > > Elapsed time: 2 hours 45 mins 33 secs > > Priority: 13 > > SD Files Written: 4,792 > > SD Bytes Written: 418,488,802,122 (418.4 GB) > > Rate: 42131.2 KB/s > > Volume name(s): WORMW-1242|WORMW-1243 > > Volume Session Id: 9 > > Volume Session Time:1542631131 > > Last Volume Bytes: 33,060,787,200 (33.06 GB) > > SD Errors: 0 > > SD termination status: OK > > Termination:*** Copying Error *** > > > > Regarding copy jobs the FD and the SD are on the same machine. > > > > we are using: > > > > Distributor ID: Ubuntu > > Description:Ubuntu 16.04.4 LTS > > Release:16.04 > > Codename: