Hope that someone can shed some light on this.
We switched to deduplication backups with NBU7 recently.
At the beginning everything was running smoothly, but now that many backup policies are running (as well as duplication to tape via SLP) we start getting error 84 far too frequently.
In the details of the job we see write errors and read errors.
Is there any kind of resource contention on the media deduplication pool that we should fine tuning?
Please see an excerpt of the bptm log on the media deduplication server:
16:34:26.375 [5408.4648] <16> media.company.com: PDVFS: [1] pdvfs_lib_log: _crStreamWrite: Error sending data: could not transmit size: software caused connection abort
16:34:26.375 [5408.4648] <16> media.company.com: PDVFS: [1] pdvfs_pwrite: CRStore 131072 bytes for mimbas41_1275740466_C1_F1.map failed: 54 (data transmission failed (software caused connection abort) (e:\pbe\darrieus-pdde\x86_64\darrieus-pdde\src\svn\libs\clibs\libcr\stream.c:soSend:1154))
16:34:26.375 [5408.4648] <16> media.company.com: write_map error (5 Input/output error) write map file entry (18446744073709551615 15376057805)
16:34:26.375 [5408.4648] <16> media.company.com: write_buf write_map failed: system call failed
16:34:26.375 [5408.4648] <16> media.company.com: write_multi_buf error exit (1485 0x1 2060017:system call failed)
16:34:26.375 [5408.4648] <16> media.company.com: write_unknown_len_buf error exit (1485 0x1 2060017:system call failed)
16:34:26.375 [5408.4648] <16> media.company.com: impl_write_image: STAT contents:
16:34:26.375 [5408.4648] <16> media.company.com: impl_write_image error exit (1485 0x1 2060017:system call failed)
16:34:26.375 [5408.4648] <32> do_write_image: sts_write_image failed: byteswritten = 1485, len = 2048
16:34:26.375 [5408.4648] <32> do_write_image: sts_write_image failed: error 2060017 (15376056320 2048 1485)
