Thank you. When the HD crashed, I took the computer to a tech. He ran all
the tests and said it was only the HD that had gone bad. He rescued most of
my data and saved it to the 2nd HD, and installed Windows Home XP on the 2nd
drive too. I've run the Windows utilities and the HD mfg's utilities and
everything comes up clean.
I'm fairly sure the troubles are USB related. For now, I'd like to know
what settings or tweaks can prevent chkdsk from restarting.
Lady D
message | On Wed, 27 Jun 2007 14:42:26 -0700, "Lady Dungeness"
|
| >Thank you -- I reviewed the articles. I installed the system 3-4 days
ago;
| >I'm trying to get it running again after a HD failure, before I install
my
| >new HD, new CD burner, and new Monitor. It ran well for 2 days, and the
| >monitor is fine. I'm not getting chkdsk errors anymore -- I canceled out
of
| >it and it hasn't reappeared. But I've started having all kinds of USB
| >troubles.
|
| Generally, primary problems (bad RAM, bad HD) can spawn secondary
| damage (file system bit-rot after bad exits, corrupted file content
| from bad RAM, lost material from bad sectors).
|
|
| The first thing to do is to check hardware, in case it wasn't only the
| HD that went bad - so I'd look at fans, motherboard capacitors and do
| a 24-hour run in MemTest86 with a different boot disk in place (so any
| spontaneous reset will be obvious, even if you aren't there to see it)
|
|
| The second thing I'd do is make sure I'm not waide open to a mass of
| ITW direct network exploits, i.e. that:
| - I have a firewall enabled
| - I am SP2 code base, or SP1 + patches, or SP0 + patches
|
| By "patches", I'm specifically thinking of RPC patch to fix the hole
| that Lovesan et al use, and LSASS patch to fix the hole that Sasser et
| al use. By June 2007, the bulk of the malware using these exploits
| will be working bots, rather than the original milder PoC worms.
|
| Also, if any HD > 137G, you cannot install SP0 (XP "Gold") and you
| should upgrade SP1 to SP2 as soon as you can. Some of SP1's code is
| not safe > 137G, and that may include the code that writes memory
| dumps to disk when the system crashes. Join the dots?
|
|
| Then I'd minimize bad exits, by killing two dumb-ass duhfaults that
| cause sick systems to automatically restart (sticking pins into voodoo
| dolls of the devs concerned is optional, but fun).
|
| The settings to change, are:
|
| Start, Settings, Control Panel, System icon, Advanced tab
| - Startup and Recovery section, Settings button
| - System Failure section, [_] Automatically restart, Apply, OK
|
| Start, Settings, Control Panel, Admin Tools, Services
| - scroll down to Remote Procedure Call (RPC)
| - Recovery tab; set all 3 drop-downs to "Restart the SERVICE", OK
|
| At least now you can see what STOP error codes you get, etc.
|
| BTW, how did you move the stricken installation from old HD to the new
| one? Most partition imaging processes may carry over "bad cluster"
| markers with the file system that aren't applicable to the new HD.
|
| >One article said that a damaged registry can cause all kinds of problems.
|
| This is true. But what would damage the registry? And more to the
| point, is this ongoing damage or just static leftovers from the
| previous bad HD?
|
| The only recent registry backups that XP creates on a cyclical basis,
| are those in C:\SVI as created by System Restore. I'd harvest those
| (all of those) from (say) a Bart CDR boot, before normal use purges
| what may be the last good ones, as new Restore Points are created.
|
| See...
|
|
http://cquirke.blogspot.com/2006/10/bart-vs-badpoolcaller.html
|
| ...for an article that walks (tersely) through the process, and...
|
|
http://cquirke.blogspot.com/2006/07/repairing-safe-mode-safeboot.html
|
| ...for more on the same sort of process. To learn more about Bart,
| you can just Google( Bart PE ) and take it from there.
|
| >It also said not to use the registry save / destroy / rebuild solution if
| >the system is OEM. I still do not know if my XP Home is OEM or not. A
shop
| >built it for me, I specified a RETAIL version of XP, they gave me a disk
| >with a holographic surface and said yes it's retail ... same as retail.
|
| You can use Nirsoft's Product Key tool to see your XP's product ID and
| product key. If the ID is xxxxxxx-OEM-xxxxxxx then it's OEM, but
| that's not always a bad thing; generic OEM is pretty much as good as
| retail, other than it won't install as an upgrade.
|
| The really sucky OEM distros are those from the big "royalty" OEMs,
| who fob users off with "Recovery Disks" or no OS disks at all.
|
| Retail is either very expensive, or crippled in other ways (i.e. the
| need for another valid OS license to upgrade from). If I had a client
| ask for the full retail pack, I'd ask "Are you sure/nuts?" ;-)
|
| >The other article explained how to run a command to find out if the C:
drive
| >is dirty. The answer is NO - C is NOT DIRTY!
|
| C: being "dirty" is just a matter of whether the OS:
| - detected a failed sector access -> triggers AutoChk like ChkDsk /R
| - detected interrupted file system updates -> AutoChk like ChkDsk /F
| - was explicitly set dirty to force one of the above tests
|
| It doesn't say anything about whether your HD and/or file system are
| OK or not, and testing file system logic won't pick up corrupted
| contents of files that were previously "fixed" or bent through bad RAM
|
| >If anybody has any other suggestions, I'd appreciate it.
|
| As above. Suggestions-R-us ;-)
|
|
|
| >------------ ----- ---- --- -- - - - -
| Our senses are our UI to reality
| >------------ ----- ---- --- -- - - - -