I have just installed Netbackup 7.5 on a Solaris 10 server using NDMP to communicate with a NetApp (running Ontap 8.1) which is connected to an Overland Neo 2000 series tape library via Sanbox 1400 (I'm waiting for a replacement tape drive so I only have one tape drive in the library right now). My basic problem is that the backups take an extreme amount of time to complete. The backups (from the Netbackup Server's perspective as seen from the activity monitor from the Admin Console) seem to start OK but the throughput is only operating at about 19-20MB/s. Most backups generally fail because the windows close before they can be addressed. When I log into the NetApp and observe the tape writes (using the command "sysstat -x 1") I noticed that the output was showing values of anywhere from around 35MB/s to 165MB/s but, after several seconds of writes it would stop for 14 - 20 seconds before beginning to write to tape again. The following is an example: (I apologize for the alignment...I hope it is readable enough to see what I'm seeing)
NetApp> sysstat -x 1
CPU NFS CIFS HTTP Total Net kB/s Disk kB/s Tape kB/s Cache Cache CP CP Disk OTHER FCP iSCSI FCP kB/s iSCSI kB/s
in out read write read write age hit time ty util in out in out
3% 1002 0 0 1002 2966 3469 2648 0 0 0 23s 86% 0% - 32% 0 0 0 0 0 0 0
3% 1106 0 0 1112 3141 2580 2492 0 0 0 23s 88% 0% - 31% 6 0 0 0 0 0 0
2% 803 0 0 803 3937 1132 752 0 0 0 23s 93% 0% - 5% 0 0 0 0 0 0 0
16% 895 0 0 895 4769 2705 66792 0 0 50397 39s 85% 0% - 58% 0 0 0 0 0 0 0
19% 976 0 0 976 2808 1951 62504 0 0 80806 39s 89% 0% - 60% 0 0 0 0 0 0 0
22% 1066 0 0 1066 3079 3116 103624 0 0 98763 25s 92% 0% - 65% 0 0 0 0 0 0 0
20% 976 0 0 983 4003 2760 78588 0 0 68944 24s 85% 0% - 66% 7 0 0 0 0 0 0
14% 1037 0 0 1038 2941 1825 33412 0 0 33554 24s 82% 0% - 56% 1 0 0 0 0 0 0
3% 1003 0 0 1003 3897 1796 1100 0 0 0 24s 86% 0% - 20% 0 0 0 0 0 0 0
9% 956 0 0 956 3295 1503 16568 5648 0 0 24s 97% 46% Tf 46% 0 0 0 0 0 0 0
6% 1057 0 0 1057 3027 2202 6724 1886 0 0 24s 89% 100% :f 37% 0 0 0 0 0 0 0
4% 875 0 0 887 2887 2936 2608 14052 0 0 24s 84% 62% : 29% 12 0 0 0 0 0 0
3% 1174 0 0 1177 3373 1843 1660 0 0 0 24s 89% 0% - 18% 3 0 0 0 0 0 0
3% 967 0 0 967 3833 3634 2640 0 0 0 24s 80% 0% - 31% 0 0 0 0 0 0 0
3% 969 0 0 969 3581 1743 1756 0 0 0 24s 90% 0% - 27% 0 0 0 0 0 0 0
3% 1101 0 0 1101 4565 2676 2424 0 0 0 24s 87% 0% - 36% 0 0 0 0 0 0 0
2% 785 0 0 793 2628 1364 1388 0 0 0 24s 90% 0% - 18% 8 0 0 0 0 0 0
2% 931 0 0 931 2438 1817 1836 0 0 0 24s 90% 0% - 23% 0 0 0 0 0 0 0
3% 984 0 0 984 3009 2223 1560 0 0 0 24s 85% 0% - 29% 0 0 0 0 0 0 0
3% 908 0 0 908 2748 2247 2488 0 0 0 24s 89% 0% - 30% 0 0 0 0 0 0 0
4% 1114 0 0 1114 5121 3814 3712 0 0 0 24s 93% 0% - 26% 0 0 0 0 0 0 0
9% 781 0 0 790 2216 2308 18016 8462 0 0 39s 96% 38% Tf 49% 9 0 0 0 0 0 0
5% 896 0 0 896 2315 1418 4248 18888 0 0 39s 91% 100% :f 39% 0 0 0 0 0 0 0
5% 897 0 0 899 4179 3082 4120 11368 0 1114 39s 79% 51% : 22% 2 0 0 0 0 0 0
23% 810 0 0 816 4347 1195 107412 0 0 103416 41s 89% 0% - 75% 6 0 0 0 0 0 0
16% 925 0 0 1002 2840 2742 53688 0 0 54002 41s 85% 0% - 63% 77 0 0 0 0 0 0
18% 999 0 0 1006 3969 2644 69408 0 0 60817 42s 87% 0% - 75% 7 0 0 0 0 0 0 30% 841 0 0 843 2619 2409 168996 0 0 165806 41s 96% 0% - 48% 2 0 0 0 0 0 0
24% 1233 0 0 1233 2909 1819 121868 0 0 118424 41s 91% 0% - 55% 0 0 0 0 0 0 0
16% 973 0 0 973 2706 2364 54840 0 0 49873 25s 83% 0% - 67% 0 0 0 0 0 0 0
6% 1036 0 0 1036 2768 1811 9832 0 0 7209 25s 84% 0% - 22% 0 0 0 0 0 0 0
4% 886 0 0 893 2863 2827 2584 0 0 0 25s 85% 0% - 27% 7 0 0 0 0 0 0
Here it will continue with this pattern of stop writing to tape for 16 - 20 seconds then continue with respectable write throughput for a single drive. Can someone please help...this is a real head banger...Can someone please give me some advise or direction to take at this point?
Thanks,
DRH