---------- X-Sun-Data-Type: text X-Sun-Data-Description: text X-Sun-Data-Name: text X-Sun-Charset: us-ascii X-Sun-Content-Lines: 51 > Recently I installed a 20 Gbyte 4mm Tape Autoloader which uses 4 standard >4mm DDS data cartridges in a tape magazine. The current backup software >package I am using is Networker for Solaris 4.0.2. on SunOS 5.3. > > The problem that I am having is with the tape capacity. I am >using 90 meter tapes. According to the documentation of the Autoloader >I should be getting aprox 5 Gbytes. > >Here are the results so far. Actual Data > Using Networker 3.368 Gbytes > Using Ufsdump 3.457 Gbytes > >I have used /dev/rmt/0n, 0un, 0cn, 0cbn, 0hcn. > >Total bytes were derived from Networker volume information and the Ufsdump tty >output and compared with physical disk Sorry took so long to SUMMARIZE this one, it has been lots of fun. I recieved lots of helpful information, but they have all lead me to the same results. Special thanks to : Phil.Schackman@seag.fingerhut.com rich@brake.demon.co.uk rwh@atmos.albany.edu blymn@awadi.com.AU martin@gea.hsr.it epl@Kodak.COM chaos!jordan.wfg.com!mam@uunet.uu.net paulo@dcc.unicamp.br First off I think most of these results are related to specifically to the Archive 4mm 20Gbyte Autoloader. The 4mm 90m tape according to SUN holds about 2 Gbytes uncompressed. The reason my results were higher is that hardware defaulted to compression reguardless of the /dev/rmt device being used. I received bugid 1181096 from SUN admitting problems with the autoloader and published capacity. I have included the bugid for those who are interested. I wish this information would have been released a little bit sooner. Anyway my results were very close to the results shown by the bugid. Michael Duncan Fingerhut Inc. Email : michael.duncan.@tti.fingerhut.com ---------- X-Sun-Data-Type: default X-Sun-Data-Description: default X-Sun-Data-Name: 1181096 X-Sun-Content-Lines: 132 X-Sun-Charset: us-ascii Bug Id: 1181096 Category: hardware Subcategory: tape Release summary: s1093 Synopsis: 4mm Autoloader is not capable of storing near suggested capacity Integrated in releases: Patch id: Description: 4mm Archive Autoloader has not shown that it can backup anywhere near its publicsized capacity. Typically the maximum that it has been capable of store has been closer to 2.9GB not 5.0GB. I have tried may different combinations of options to ufsdump while attempting to backup a 3.1GB filesystem. This file- system is mostly text data which will compress with unix "compress" to 1.966GB . Filesystem is /slap/support. Using dd with a 12MB textfile I was able to store a max of 3.6GB to the device, this file compressed to 6.29MB. Though it is obvious that the device is capable of storing a reasonable amount of data,considering the type of files being used, in the customer's viewpoint is the device is falling far short of its expected capacity. testcase: Files are on system flap, /corefiles/mcroke tape.sh: #! /bin/sh # echodo() { echo "+ $*" $* return $? } DEV=/dev/rmt/2bn LOOPCOUNT=1 LOGFILE="test.log" TESTFILE="testfile-cpio" ( echo "----Starting Test --" echodo mt -f $DEVrewind while echodo dd if=$TESTFILE of=$DEV bs=112k ; do echo "$LOOPCOUNT done." LOOPCOUNT=` expr $LOOPCOUNT + 1` done ) > $LOGFILE 2>&1 tail of log files: test.log_cpio.bn.112k 109+1 records out 305 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2bn bs=112k 109+1 records in 109+1 records out 306 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2bn bs=112k dd: unexpected short write - wrote only 65534 bytes, expected 114688 1+0 records in 1+0 records out ---------------------------------------- test.log_cpio.cn.112k 109+1 records out 305 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2cn bs=112k 109+1 records in 109+1 records out 306 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2cn bs=112k dd: unexpected short write - wrote only 65534 bytes, expected 114688 1+0 records in 1+0 records out ---------------------------------------- test.log_cpio.bn.56k 218+1 records out 305 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2bn bs=56k 218+1 records in 218+1 records out 306 done. + dd if=/corefiles/mcroke/testfile-cpio of=/dev/rmt/2bn bs=56k dd: unexpected short write - wrote only 0 bytes, expected 57344 2+0 records in 2+0 records out ---------------------------------------- ufsdump info # /usr/ucb/du -s /support 3127676 /support # df -k /support Filesystem kbytes used avail capacity Mounted on /dev/md/dsk/d15 9758338 3127676 5654832 36% /support # ufsdump 0ucbf 56 flap-lab:/dev/rmt/2c /support DUMP: Date of this level 0 dump: Wed Oct 26 07:44:30 1994 DUMP: Date of last level 0 dump: the epoch DUMP: Dumping /dev/md/rdsk/d15 (/support) to /dev/rmt/2c on host flap-lab DUMP: mapping (Pass I) [regular files] DUMP: mapping (Pass II) [directories] DUMP: estimated 6353996 blocks (3102.54MB) DUMP: Writing 28 Kilobyte records DUMP: dumping (Pass III) [directories] DUMP: dumping (Pass IV) [regular files] DUMP: 2.68% done, finished in 3:01 DUMP: 6.27% done, finished in 2:29 DUMP: 10.47% done, finished in 2:08 DUMP: 15.62% done, finished in 1:48 DUMP: 20.50% done, finished in 1:36 DUMP: 24.46% done, finished in 1:32 DUMP: 28.14% done, finished in 1:29 DUMP: 32.58% done, finished in 1:22 DUMP: 37.19% done, finished in 1:15 DUMP: 41.72% done, finished in 1:09 DUMP: 46.29% done, finished in 1:03 DUMP: 50.99% done, finished in 0:57 DUMP: 55.95% done, finished in 0:51 DUMP: 60.77% done, finished in 0:45 DUMP: 65.27% done, finished in 0:39 DUMP: 69.61% done, finished in 0:34 DUMP: 73.94% done, finished in 0:29 DUMP: 78.60% done, finished in 0:24 DUMP: 83.41% done, finished in 0:18 DUMP: 88.09% done, finished in 0:13 DUMP: 93.12% done, finished in 0:07 DUMP: End-of-tape detected DUMP: Tape rewinding DUMP: 94.95% done, finished in 0:05 DUMP: Change Volumes: Mount volume #2 DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted and ready to go?: ("yes" or "no") DUMP: The ENTIRE dump is aborted. Summary: 4mm Autoloader is not capable of reaching its execpted capacity using ufsdump History: Submitter: michael.croke@east Date: 10/28/94 Dispatch Operator: bugtraq Date: 10/28/94