Groups	Erweiterte Groups-Suche

Hello NG,

We recently got a QTof (and soon there would be two) and we are
generating 2 GB of data a night.  Do anybody have any thoughts on
backups the data we are generating??

CD-R vs. Tapes vs. DVD-RAM? vs. ??
Any Good archive programs?
How are you dealing with the flood of data??

We were thinking of CD-R (Cause it is so inexpensive) but at the end of
the year we could end up with 1000 CD!  Please share your experience
with the NG.

Swimming in CDs

Ken

You ask a good question.  We have a ToF-SIMS that can generate what are
called "*.raw" files that can be up to 2 Gb in size.  We've decided to
upload the date onto a server where 3 - 6 months of data can be made
availabe for more or less instant access and the post 3 - 6 month data
gets stored on tape in case it's ever needed.

Even with these large files we certainly don't normally generate 2
Gb/day.  We looked at the idea of CR-R, tapes, optical drives, Jazz
drives plus a few others and as far as we could see there isn't a good
solution to the problem.  Tapes are big on storage but too slow, most
everything else is fairly fast but too small.



Kenneth Chan wrote:
}
} Hello NG,
}
} We recently got a QTof (and soon there would be two) and we are
} generating 2 GB of data a night.  Do anybody have any thoughts on
} backups the data we are generating??
}
} CD-R vs. Tapes vs. DVD-RAM? vs. ??
} Any Good archive programs?
} How are you dealing with the flood of data??
}
} We were thinking of CD-R (Cause it is so inexpensive) but at the end of
} the year we could end up with 1000 CD!  Please share your experience
} with the NG.
}
} Swimming in CDs
}
} Ken

}We recently got a QTof (and soon there would be two) and we are
}generating 2 GB of data a night.
}
}We were thinking of CD-R (Cause it is so inexpensive) but at the end of
}the year we could end up with 1000 CD!  Please share your experience
}with the NG.
------------------------------------------------------------------------------------

Hi,
this is a very interesting discussion.

(1000 CDs are 650 MB * 1000 = 650 GByte) at a rough guess

If you need your data only in the next 5 years - take CDs  1000 * $1 =
$1000
(I have old CDs and can not read them) - but use some good quality CDs.

If you need your data in the next 5 years and want to get rid of 1000 CDs
take the new ADR streamers they are fast and have 50GB space
take the OnStream Tape-Drive SC50 50GB SCSI-2
its $600 in germany and tapes (1x =$50) will cost you again $600
http://www.onstream.com/server/pricelist.html


If you _really_ need your data only use MOs
http://fujitsu.dynamit.net/web/mo_disk/htm/index.htm
but you will get poor. Because the GIGAMO-drive costs you
around $600 but the media (1,3Gbyte 1x=$40) will cost you $20.000 (oooops)

ok buy the HP-MO solution
http://www.hp-optical.com/products/fs_products.htm
its the HP 5200ex $2500 and Media (5Gbyte = $80) = $10.000 (ooops)

If you want DVD use toshiba or panasonic DVD-RAM
but Medium 5 GByte will cost (1x=$40) you $5200 (oops)
http://www.toshiba.com/taecdpd/products/features/SDW1111-Over.shtml

Prices may differ, but best thing:
A) life time --> go for MO
B) Optimum  --> go for 50 GByte ADR-Streamer


(GC files are in most cases compressed, but you can crunch them down
to 60% of the original size)

For compression UNIX go for a ZIP derivate
For WINDOWS go for ARJ
at the and use an automatic batch converter for compressing all the files
and save them.

Kind regards
Tobias Kind

...mmhh print it out ???

Here is something we been playing around / thinking about  here:

We are going to try to setup a file server with the biggest hard disk we could find (in
Canada Maxtor Diamond Max 60 = $440 cdn.)  When the hard drive gets full we will pull
the full hard drive out and replace it with a fresh HD.

Cost per Gb ( in Canadian dollars)
HD = $7.33
DVD = $12.50 ($65 / disk 5.2Gb)
Tape = $3.00 depending on formats
CD = $2.00

It sounds crazy, but it has the fastest transfer rate of the medium and it does not
require an interface.

Got any better ideas???  Cause this sounds too crazy to work (or could it?)

Ken

Tobias Kind wrote:

} Hi,
} this is a very interesting discussion.
}
} ---> sorry for the snip!
}
} If you want DVD use toshiba or panasonic DVD-RAM
} but Medium 5 GByte will cost (1x=$40) you $5200 (oops)
} http://www.toshiba.com/taecdpd/products/features/SDW1111-Over.shtml
}
} Prices may differ, but best thing:
} A) life time --> go for MO
} B) Optimum  --> go for 50 GByte ADR-Streamer
}
} Kind regards
} Tobias Kind
}
} ...mmhh print it out ???

All,

In regards to data acquired on Q-TOF type instruments.  Although
expensive, the data capacity of a RAID type 3 or type 5 would be much more
benificial. for $5,000 (US) you can now purchase units up to 250Gb (5x50Gb
drives)  They all sit in a single case, and then the processed data can be
backed up onto another media (i.e. CD-R)  this would provide fast access
to the raw data, and then allow for selection of the archival data.

This is the method now being used for DVD development of Video.

Regards,

Ken Weir
Media Development Coordinator
MDS Sciex Instruments
Kenneth Chan <Kenneth.Chan@nrc.ca> wrote in message
news:8lkheq$9ks$1@news-int.gatech.edu...
} Here is something we been playing around / thinking about  here:
}
} We are going to try to setup a file server with the biggest hard disk we could find (in
} Canada Maxtor Diamond Max 60 = $440 cdn.)  When the hard drive gets full we will pull
} the full hard drive out and replace it with a fresh HD.
}
} Cost per Gb ( in Canadian dollars)
} HD = $7.33
} DVD = $12.50 ($65 / disk 5.2Gb)
} Tape = $3.00 depending on formats
} CD = $2.00
}
} It sounds crazy, but it has the fastest transfer rate of the medium and it does not
} require an interface.
}
} Got any better ideas???  Cause this sounds too crazy to work (or could it?)
}
} Ken
}
} Tobias Kind wrote:
}
} } Hi,
} } this is a very interesting discussion.
} }
} } ---> sorry for the snip!
} }
} } If you want DVD use toshiba or panasonic DVD-RAM
} } but Medium 5 GByte will cost (1x=$40) you $5200 (oops)
} } http://www.toshiba.com/taecdpd/products/features/SDW1111-Over.shtml
} }
} } Prices may differ, but best thing:
} } A) life time --> go for MO
} } B) Optimum  --> go for 50 GByte ADR-Streamer
} }
} } Kind regards
} } Tobias Kind
} }
} } ...mmhh print it out ???
}
}
}

Hi Ken,

Do you have an idea what and especially WHEN you will do something (e.g.
process) with this data? This will maybe lead you to your appropriate
solution. If you don't get the time and capacity to process you can as
well record 2 GB of noise a night.

IMHO the solution is to process the data online, i.e. as fast as you
generate them. Archive only the REAL RESULTS.
If you just store the masses (and maybe a width and a reliability
score), you will have only one CD instead of 1000.

Jens


Kenneth Chan wrote:
}
} Hello NG,
}
} We recently got a QTof (and soon there would be two) and we are
} generating 2 GB of data a night.  Do anybody have any thoughts on
} backups the data we are generating??
}
} CD-R vs. Tapes vs. DVD-RAM? vs. ??
} Any Good archive programs?
} How are you dealing with the flood of data??
}
} We were thinking of CD-R (Cause it is so inexpensive) but at the end of
} the year we could end up with 1000 CD!  Please share your experience
} with the NG.
}
} Swimming in CDs
}
} Ken

Hello Jens

We do "Real Time" processing (i.e.. When the MS/MS spectrums are ran, we do
an automated peptide search).  And yes most of the time we only look at the
processed data.  However, deleting the original data file is not possible
because of chain of causticity.(sp?) issues, i.e.. where the data comes form,
how it was processed, can we get anything else out of the data?

The scary part is that we have rarely had 2 GB of junk...  Only twice did an
over night run gone bad.  And in was an HPLC problem, not the mass spec....

Ken

Jens Höhndorf wrote:

} Hi Ken,
}
} Do you have an idea what and especially WHEN you will do something (e.g.
} process) with this data? This will maybe lead you to your appropriate
} solution. If you don't get the time and capacity to process you can as
} well record 2 GB of noise a night.
}
} IMHO the solution is to process the data online, i.e. as fast as you
} generate them. Archive only the REAL RESULTS.
} If you just store the masses (and maybe a width and a reliability
} score), you will have only one CD instead of 1000.
}
} Jens