romeo@lindy.stanford.edu (Patrick Goebel) (06/02/90)
This is in reponse to Danielle's question about backing up to 150-MB tapes instead of the 60-MB ones. These are the parameters I was told to use by our Sun FE: /usr/etc/dump [0-9]uctsdf 18 620 1250 /dev/nrst0 /dev/rsd6[a-h] where the named devices are obviously specific to your system and the tape length (in this case 620') might vary according to manufacturer. Using these parameters, dump issues reasonable estimates of needed tape capacity for a given file system, assuming the full tape capacity is 150 MB. Patrick Goebel -- romeo@lindy.Stanford.EDU
gary@sunkist.west.sun.com (Gary Cattelino) (06/07/90)
> This is in reponse to Danielle's question about backing up to 150-MB tapes > instead of the 60-MB ones. These are the parameters I was told to use by > our Sun FE: > /usr/etc/dump [0-9]uctsdf 18 620 1250 /dev/nrst0 /dev/rsd6[a-h] On page 74 of the Sparcstation 1 Sun System and Network Manager's Guide (the little books you get when you buy a Sparc 1) it says to use: /usr/etc/dump 0cusdtf 590 1200 18 /dev/nrst0 filesystem In the discussion they say to use 590 to handle media errors (600 feet minus 10 feet to handle media flaws). Well, I just looked at a DC6150 and it says 620 feet. Maybe the proper size to use for these tapes is 610 feet? It also says the density is 1200, not 1250 from the previous posting. Which is correct? I've been using 1200 with no problems. Maybe a density of 1250 will allow me to store more on one tape? Anyone know the answers to these questions? Gary A. Cattelino Unocal Science & Technology Division Phone: (714) 528-7201 x2715 376 S. Valencia Avenue UUCP: uunet!zardoz!cerveza!gary Brea, CA 92621 sun!sunkist!cerveza!gary INTERNET: gary@unocal.COM
del@thrush.mlb.semi.harris.com (Don Lewis) (06/08/90)
In article <8624@brazos.Rice.edu> cerveza!gary@sunkist.west.sun.com (Gary Cattelino) writes: >> This is in reponse to Danielle's question about backing up to 150-MB tapes >> instead of the 60-MB ones. These are the parameters I was told to use by >> our Sun FE: > >> /usr/etc/dump [0-9]uctsdf 18 620 1250 /dev/nrst0 /dev/rsd6[a-h] > >On page 74 of the Sparcstation 1 Sun System and Network Manager's Guide >(the little books you get when you buy a Sparc 1) it says to use: > >/usr/etc/dump 0cusdtf 590 1200 18 /dev/nrst0 filesystem > >In the discussion they say to use 590 to handle media errors (600 feet >minus 10 feet to handle media flaws). Well, I just looked at a DC6150 and >it says 620 feet. Maybe the proper size to use for these tapes is 610 >feet? It also says the density is 1200, not 1250 from the previous >posting. Which is correct? I've been using 1200 with no problems. Maybe >a density of 1250 will allow me to store more on one tape? > >Anyone know the answers to these questions? Disclaimer, I don't have any authoritative info. A couple of weeks ago I did some capacity calculations and figured that if the density is 1250, then the capacity works out about right. Also, I just noticed that the default density with the "c" option is 1000 and DC300XL/P tapes are rated at 10000 ftpi, and that DC6150 tapes are rated at 12500 ftpi. It sounds like the density should be 1250. I have the following tapes in front of me: 3M DC6150 600 feet (182.9m) Carlisle DC600XTD (Series II Gold Plus) 620 feet (189.0m) DC600XTD is supposed to be the same as DC6150 DEI Series II Gold (appears to be DC600A) 600 feet (182.9m) 3M DC600A 600 feet (182.9m) 3M DC600A 620 feet (189.0m) So much for standardization. It looks to me like the proper parameters are: length 590/610 (depending on the tape) density 1250 By the way. There is supposed to be a QIC-120 tape format which holds 125 Meg on a tape. The difference comes from putting only 15 tracks on the tape instead of 18. I tried writing some DC600A tapes on one of the Wangtek drives here and it seems to hit EOT at around 130 Meg, as opposed to 155 for a DC6150. According to the labels on the tapes, the lengths, magnetic properties, and rated density are the same. Apparently the drive can sense the tape type and do the appropriate thing. I have not yet tried this on an official Sun (Archive?) drive. Has anyone heard if this is actually supposed to work? Don "Truck" Lewis Internet: del@mlb.semi.harris.com
grenvill@trajan.uk.sun.com (Grenville Whelan - Sun UK) (06/11/90)
In article <8624@brazos.Rice.edu> cerveza!gary@sunkist.west.sun.com (Gary Cattelino) writes: >X-Sun-Spots-Digest: Volume 9, Issue 202, message 4 > >In the discussion they say to use 590 to handle media errors (600 feet >minus 10 feet to handle media flaws). Well, I just looked at a DC6150 and >it says 620 feet. Maybe the proper size to use for these tapes is 610 >feet? It also says the density is 1200, not 1250 from the previous >posting. Which is correct? I've been using 1200 with no problems. Maybe >a density of 1250 will allow me to store more on one tape? > >Anyone know the answers to these questions? For QIC cartridges, I don't use the "-c" option on dump, but calculate all parameters as follows :- Density for QIC always 1000bpi Tape Length is real length * number of tracks * 0.9 to leviate 10% for errors, Therefore, for a 60Meg (QIC24/9TRK) tape, density is 1000bpi, and length is (600 * 9 * 0.9), = 4860. eg. dump 0usfd 4860 /dev/rst8 1000 /dev/sd.... For a 150Meg (QIC150/24TRK) tape, density is 1000bpi, and length is (600 * 24 * 0.9), = 12960. eg. dump 0usfd 12960 /dev/rst8 1000 /dev/sd... Hope this helps, Grenville Grenville Whelan, Sun UK, Hilton House, Lord St, Stockport, Cheshire, SK1 3ND [Tel. +44 061-477-3100, Email: gwhelan@sunuk.Sun.COM] "I'd rather have a full bottle in front of me than a full frontal lobotomy!"