All, I'm troubleshooting an issue with a server where
full backups succeed but the incrementals tend to fail.
Master/Media: Windows 2003 running NBU 6.0
MP4
Client: Windows 2000 running NBU Client 5.1
MP6
The client's bpbkar log has tons of errors like the
first line below for just less than an hour before the backup dies. I've
got this system backing up each drive in a separate job and most drives succeed,
but the one with all the data (and little files) consistently fails.
Occasionally the OS and a few smaller partitions fail with error 41 as
well.
4:52:44.362 PM: [2636.440] <4> dos_backup::tfs_include: INF
- folder (2008-03-27) has been created recently (since 3/21/2008 7:12:06
PM). It will be backed up in full. 5:02:22.498 PM: [2636.440]
<16> dtcp_write: TCP - failure: send socket (720) (TCP 10054: Connection
reset by peer) 5:02:22.498 PM: [2636.440] <16> dtcp_write: TCP -
failure: attempted to send 108 bytes 5:02:22.498 PM: [2636.440] <2>
tar_base::V_vTarMsgW: FTL - catalog write error (1) 5:02:22.498 PM:
[2636.440] <16> dtcp_write: TCP - failure: send socket (720) (TCP 10054:
Connection reset by peer) 5:02:22.498 PM: [2636.440] <16> dtcp_write:
TCP - failure: attempted to send 6 bytes 5:02:22.498 PM: [2636.440] <4>
ov_log::OVLoop: Timestamp 5:02:22.514 PM: [2636.440] <2>
tar_base::V_vTarMsgW: INF - EXIT STATUS 14: file write failed 5:02:22.514 PM:
[2636.440] <16> dtcp_write: TCP - failure: send socket (720) (TCP 10054:
Connection reset by peer) 5:02:22.514 PM: [2636.440] <16> dtcp_write:
TCP - failure: attempted to send 6 bytes 5:02:22.514 PM: [2636.440] <4>
tar_backup::backup_done_state: INF - Not waiting for server
status 5:02:22.514 PM: [2636.440] <16> dtcp_write: TCP - failure: send
socket (720) (TCP 10054: Connection reset by peer) 5:02:22.514 PM: [2636.440]
<16> dtcp_write: TCP - failure: attempted to send 6 bytes 5:02:22.514
PM: [2636.440] <4> OVStopCmd: INF - EXIT - status =
0
Any ideas what the dos_backup::tfs_include line is about? Its frustratng troubleshooting because the error dosen't happen every time.
-Jonathan
On Fri, Mar 28, 2008 at 3:52 AM, Dariusz Klar < Dariusz.Klar AT sun DOT com> wrote:
Recently
I noticed that indexes are consuming a lot of disk space. Does anyone know
the rule, how to plan space for indexes? Looks like something has changed
since older versions of NetBackup.
I don't think anything has changed. The challenge in planning is that
the catalog size is based on number of files you're backing up, not the size of
the files. If you back up a lot of small files, your catalog will be much
larger than the site that backs up the same amount of disk space containint
large databases. The longer the retentions, the more likely it is that
you'll need a larger catalog for backing up the same amount of disk space.
The golden rule to remember is to never, ever fill your catalog disk.
Use a different file system for your logs since they're the most likely thing to
fill the disk.
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|