Language:
switch to room list switch to menu My folders
Go to page: First ... 25 26 27 28 [29] 30 31 32 33 ... Last
[#] Fri May 04 2007 08:51:18 EDT from Ian M. Shot @ Haven BBS

[Reply] [ReplyQuoted] [Headers] [Print]

It isn't just Databases. Documents, e-mail, other data that my users create, and all that stuff.

 

However for them we could, if we had another site to replicate to, and enough bandwidth between said mythical sites and time to do so.  However since we have only one location replication would be fine if I had the spare hardware also to duplicate the server.   Since I have 3 different DB servers, each running a different type of DB (MySQL, MS SQL, and Oracle), it would take more expertise than I have to replicate between them for meaningful backups.  So tape is a bit easier in this instance. 

Thu 03 May 2007 04:19:31 PM EDT from wizard of aahz@uncnsrd (Uncensored)

Ian - Just out of curiosity, if it's databases needing backup why not use replication?



[#] Fri May 04 2007 09:28:32 EDT from IGnatius T Foobar @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Back at the low end, I've been giving some consideration to putting an alternate group of disks online at home for multimedia (video, audio, and photos). Disk is cheap; I'm more concerned with the cost of the electricity for running all those disks. Does anyone know of NAS or SAN appliances that include the ability to spin down the disks when not in use? I certainly don't mind having to wait for disks to spin up when first accessed.

[#] Fri May 04 2007 11:34:47 EDT from Spell Binder @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

IG: Would a software RAID solution or a cheap hardware RAID card do the trick?
If so, couldn't you just use your OS' power-management software to spin the disks down?

[#] Fri May 04 2007 12:27:08 EDT from IGnatius T Foobar @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

I know how to do that on IDE (sorry ... ATA) drives in Linux. You use something like:

hdparm -S 1 /dev/hda

How do you do that to a SCSI, SATA, or USB-attached drive?

[#] Fri May 04 2007 15:31:43 EDT from athos-mn @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

We're going to try out a new solution where I work, it's a separate server (appliance, really) that does incremental backups every 15 minutes by partition, and if necessary uses the backup data to launch a virtual copy of the server. Should the building get whacked by a meteor, there's an online copy stored elsewhere, which can be transfered to another appliance and flown in overnight.

 

Probably not practical for some of the larger amounts you've been talking about, but nice for a good chunk of them. 



[#] Mon May 07 2007 17:09:49 EDT from Spell Binder @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

IG: Re: Powering down hard drives in Linux.

I'm not well-versed in Linux, so I'm not completely sure how to power down SCSI or SATA drives, but a little Googling came up with a couple things to look at.

Could you enable APM support and use that to power-down the drives? If I understand it correctly, APM is a function of the BIOS, so as long as Linux can tell APM what to do, it should be able to power things down.

I did find reference to a FreeBSD utility called "camcontrol" which can power down or up hard drives. I don't know if there's a Linux equivalent, though.
Power Binder

[#] Thu May 10 2007 15:43:36 EDT from Spell Binder @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Would drives that implement SMART fall into that category?

[#] Thu May 10 2007 15:46:13 EDT from Spell Binder @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Ford: Re: Programmable hard drives.

It's not an entirely new concept. IBM's 370-based mainframes had I/O channels, which were essentially smaller-scale programmable CPUs designed to deal with specific I/O modules. The main CPU would download a program onto the I/O channel for a specific device to help off-load processing.

Also, didn't Commodore's C=64 and C=128 have programmable floppy drives?
Channel Binder

[#] Thu May 10 2007 15:57:26 EDT from fleeb @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


Thu 10 May 2007 03:46:13 PM EDT from Spell Binder@uncnsrd

Ford: Re: Programmable hard drives.

Also, didn't Commodore's C=64 and C=128 have programmable floppy drives?
Channel Binder

I thought it was Apple that had th programmable floppy drives. You could even tell them how fast to spin the drive (and people used to create their own variable-speed floppies to create a kind of copy protection).

[#] Thu May 10 2007 16:09:16 EDT from IGnatius T Foobar @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

It was Commodore's floppy drives (such as the 1541 and 1571) that were essentially built as self-contained computers. Each drive had its own 6502 CPU and 2KB of memory. There was a 16K ROM which contained the code that implemented the filesystem and disk utilities.

Since the drives were completely programmable, you could write additional code of your own. I once built a tiny record manager that ran *on* the drive, and did its own indexing and sorting on the drive. This type of thing was quite common because the speed of the serial bus back to the host computer really sucked at the time, so you offloaded whatever disk-intensive operations you could.

[#] Tue May 15 2007 17:18:01 EDT from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Also, didn't Commodore's C=64 and C=128 have programmable floppy
drives?

that's the bit of history I was alluding to, I didn't know if anybody would pick it up.

[#] Tue May 15 2007 17:19:01 EDT from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

yea, the apple drive was about as dumb as it got.
You could 'program' it in that you can pretty good control of things like head position and on off. You couldn't control speed though, not on the disk ][.

[#] Tue May 15 2007 17:23:11 EDT from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


It was horribly dumb. If I remeber correctly, you had direct control of the track-track stepper motor (you had to time when it turned on and off, and you could turn it off early to do certain copy protection schemes involving half-tracks or spiral tracks) and to cause data to be written out of the drive head, you had to write to a register that was pretty much directly connected to the write head. So you had to time the rate that bits were written out of the head pretty carefully. The whole mess was handled on the CPU, and in a mix of software and firmware.

[#] Tue May 15 2007 17:24:39 EDT from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

like head position and on off. You couldn't control speed though, not

on the disk ][.

Not in software, but there was a pot that you could adjust in the hardware to tune the spin speed.. This was actually necessary to copy certain protected floppies.

[#] Wed May 16 2007 08:43:07 EDT from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

the rate that bits were written out of the head pretty carefully. The

whole mess was handled on the CPU, and in a mix of software and
firmware.

yea, wasn't it great?!?!?!

[#] Wed May 16 2007 08:44:22 EDT from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

well, yea, I was referring to software.
I also just remembered that you could stop that extra second spindown time of the drive because I remember it would only power one drive at a time so if you wanted to turn off drive 1 you power drive two and 1 would immedately stop.

[#] Fri May 18 2007 20:11:34 EDT from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


ooh now some apple trivia I didn't know. If you start getting into ProDOS internals I'll only have vague recollections on tap to compete...

[#] Sat May 19 2007 23:10:15 EDT from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Let's see what do I remember about prodos.
It had a hierarchical tree node leaf file system which only scaled so much and I think got as inefficient as FAT once it got too big, but hard drives weren't that big back then.
Looking back, it's pretty amazing they spent all that time and money on a new DOS knowing the thing was on its way out.
And it was a really well designed DOS as things went back then, and it was all in assembly...

[#] Sun May 20 2007 00:04:58 EDT from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


Yeah, from what I could tell of it, it seemed really elegant. Though it got in the way sometimes.

I had a disk full of notes on the internals that I got from some shareware compilation vendor...

[#] Tue Jun 05 2007 08:16:08 EDT from Ragnar Danneskjold @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Emacs 22 is out. For all of you who have been patiently waiting the last 6 years for it, you're idiots.

Go to page: First ... 25 26 27 28 [29] 30 31 32 33 ... Last