OT: redhat

[Bill Vermillion]

Shakespeare wrote plays and sonnets that will last an eternity, 
but on Sun, Nov 07 18:04 , Fairlight wrote:" 

> You'll never BELIEVE what Bill Vermillion said here...:

> > I've seen that too. Using 2GB+ is not unusual in Linux, but using 
> > more than 100MB in FreeBSD / is unusual.  Though the 5.x has

> I think Bill said that one of the latest (SuSE 9.2??) uses over
> 6 gigs if you install everything, wasn't it?

> > much larger defaults.  Now that 5.3 is officially released and
> > 6.0 is now about two steps past the bleeding edge, well start
> > seeing more people migrating to the 5.x.

> What happened to 5.[012]?

5.0 came out in January 2003.  It was the first of CURRENT - which
is always the bleeding edge and that name leads a few astray.

5.1 came out in June of 2003.

5.2 in January of 2004

5.2.1 in February of 2004.

The the 5.3RC [Release candidates] started coming out and minor
things were released, and the official RELEASE - 5.3 - was
made on Novemeber 6.   All 5.x up to RELEASE were CURRENT.

Now that 5.3 is in RELEASE the CURRENT becomes 6.0.  That means
that STABLE is now 5.3.  The difference between RELEASE and STABLE
is that STABLE can change daily, but no changes are made to RELEASE
unless there is a security patch.

I guess you hadn't been noticing when others put in earlier 5.x -
and I think Brian tried it well over a year ago.

> > I suspect that is going to catch a few people unaware who have
> > applications that depend on file system structures and expect
> > such things as the long standing 128byte inode size.  And with the
> > large inodes we have a create time for the first time.

> That affect fP in any way?

You'll have to ask Ken about that.   But that is a major change for
any program that looks to the lower levels.  Remember when the
calcuations to see if there was disk space left broke when OSes
were able to break the 2GB filesystem limit?

UFS2 is the first truly major change to the UFS file system in
20 years.  UFS2 installs by default on the 5.3, but you can
choose the standard UFS.

> > SCO mis-documented ctime as innode creation time instead of CHANGE
> > time, and that mis-understanding has propagated widely.  So
> > now there is access time, modifciation time, change time, and
> > creation time.   It also means that we can have much larger files

> What is it, st_rtime?

It's a little more understandable than that :-)   The following
are from stat.
st_atime - file access time
st_mtime - file modification time
st_ctime - file status change time (inode data modification time)

and the new one is

st_birthtime - time when inode was created

:-)

> > I >>HATE<< it when backups are stale and you have to mount
> > / in read-only mode and then slowly recover things a bit at a time.
> > Worst was years ago when the lost+found was not large enough to be
> > able save things, and answering Y to remove things would leave a
> > large data loss.  So that was a mount read-only, save files
> > externally, and do it over and over.

> Talk about stale...I once had to take a call from a client of
> a client who had their console monitor go dead, nobody ever
> checked root's email, and nobody ever saw their tape backups
> weren't succeeding -or- verifying.

And that's why I always have BackupEdge or LoneTar send me emailand
mail to the client.  Depending on the client to do that is not
good.  One time when I called a client I hadn't heard from in six
months the woman who was now taking care of the daily routine had
never been told to check the email.  Their tape drive had failed
sometime in the past 3 to 4 months.

> They religiously changed tapes, but only the lady that was on
> the phone with me that day seemed to notice that at least three
> of the tapes had actually come off their spools entirely. NONE
> of the backup tapes worked--at all. They had to get their DBIV
> programmer to recover what he could from what was left. I can't
> remember exactly, but I think the fragment that was left was
> about 27MB of a file that should have been over half a gig.

I recovered a system where the tape had come off.  After
painstaking reloading the tape and spending a few hours
reconstructing their system I got them back up and running.

They never paid the bill though I saved them well in excess of
$25,000 in losses.   I should have learned well enough from that
client while doing some work for him before getting in computers
full time.

> They had Edge--they just never paid attention to the
> notifications, never replaced the monitor because nobody ever
> used the console anyway, and for whatever reason, root's mail
> was never redirected -or- read.

I went into one place that called me occaisionally because the
people who maintained thier SW weren't around.  So the first thing
I did was make a backup of the system in case I screwed up - having
never seen the SW before.  Then I checked the tape for readability.

No go!  I guess that all their daily backups for the past 6 months
were bad.  I have never seen a tape drive so dirty.  I went through
at least a dozen swabs before they stopped coming out completely
black. And considering how debris buildup on a head can act like
a razor and each pass makes the tape worse I had them buy a
complete new set of tapes too.

[An aside.  The old reel-reel 9-track tapes would often be
reconditioned by running them through a device that acted similar
to a carpenters plane and made the tape surface smooth.  Drop outs
are caused by particles on the surface so the tape is basically
made smooth again. In the early audio days before Irish invented
the calendaring process [which caused Ampex to buy them to get
that process] if you wanted to be SURE your recording was going to
turn out right you ran the tape through the machine at least once
to knock off all the high sposts.  Between broadcast, recording,
computers and hobby, I've learned more about mag tape than I ever
suspected was possible].

> > I got an email one day with a 1/2 dozen images.  Smoke and flames
> > rolling out of the building.  NOTHING was salvageable.  And his
> > only backup of the Linux filePro was made from an MS machine that
> > truncated all names that did not fit in 8.3 name-space.

> Gack!

> > Quoted a lower than normal hourly rate and figured it would
> > be at least $5000 to get it back to the state it was in the month
> > before the fire.  In the end I think he wrote it all off, and
> > actually focused on building his other businesses.

> On the whole, probably wisest for all concerned. You'd have
> gotten screwed in all likelihood. You know what a time-sink
> that stuff can be, especially when they barely remember what
> they had in some places.

Not this one.  That company was one he sold when he got into what
turned out to be a dot-bomb business and he took it back when the
other people failed making payments.  It had FP code I had moved
from a Model III back in 1984 through several machines.  He's now
gotten out of having a business where he needs to be near freight
handlers and is now doing brokering and referencing on line.
His site kept getting busier and busier until he now has his
own dedicated rack mount machine in our facility.  Running MySQL
and about 10 different sites.

> > Haveing a BE/LT daily backup to tape with at least one a week going
> > off site would have meant he would have been operational in less
> > then a day.

> People really should replace their media once in a while, as
> well. I can't believe anyone could -miss- tapes being off the
> reels. But tape just goes bad after a while, as well. I dunno
> what the rule of thumb is, but I should guess that replacing
> the batch of tapes at least once a year is a prudent move.

Those had to be the old QIC tapes.  Since the tape is basically
carried on a band [sort of like an open face sandwich] it is easy
to think the tape is OK.  You have to spring open the window to
check.  And if you don't keep the system at least somewhat clean
you can get dust/dirt built up either on the mirror or the sensor.

When that happens the tape can't read the holes and goes right 
off the spool.  That had happened on the above site that I didn't
get paid for.  I had to go in and clean that area too.  [There are
times that I wish I didn't know both hardware and software.
Particuarly when people start putting the machine in inaccesible
locations].

Bill

Bill
-- 
Bill Vermillion - bv @ wjv . com