Help with a Dimension 8200 Memory and Freezing
in [Dell]
|
From: RnR on 8 Feb 2010 19:46 On Mon, 8 Feb 2010 18:34:21 -0600, "William R. Walsh" <newsgroups1(a)idontwantjunqueemail.walshcomptech.com> wrote: >Hi! > >> Has SMART improved because from my past experience, SMART info was >> wrong? > >Yes and no. > >Every drive maker has a different idea about how SMART should be >implemented, how errors should be detected and what categories they should >watch. And the thresholds all differ as well--which is to say that each >manufacturer has a different idea of "how bad" things have to get before the >drive is considered to be failing. > >Out of all of the drives I've seen, I'd say that Maxtor has the most >informative SMART reporting. Yes, believe it or not. Their drives monitor a >wide variety of parameters and the values do shift around while the drive >operates. Second in line would be Seagate, whose drives don't show as many >parameters but still seem to be pretty honest. Quantum drives fit somewhere >in between, and Quantum was actually one of the first companies to implement >SMART in a hard drive. > >Hitachi GST/IBM drives have a similar selection of monitored parameters, but >I'll have to admit that I've never seen them vary a whole lot, even on very >obviously sick drives. ExcelStor drives behave very similarly, which isn't >too surprising considering that A) their drives are mostly Deskstar clones, >B) HGST made some of their control boards and C) someone told me that HGST >owns ExcelStor now. > >Western Digital (who makes otherwise excellent hard drives in my experience) >has typically been very conservative with SMART data...with only a few >monitored parameters and seemingly little variation, even on drives that >were starting to have problems. > >Of course, SMART data is sometimes only collected by a drive when it is >idle. In a modern operating system, between power management events (spin >down) and disk activity in general, a drive may never get a chance to do its >SMART self test routines because it either isn't idle for long enough, or >it's spun down. (Other drives seem to observe what's going on while they are >active, such as drives from Seagate, HGST and Maxtor.) So sometimes a drive >needs a little prodding to update its SMART data and notice a problem. That >kind of defeats the purpose. Tools such as SpinRite, HDAT2 and SpeedFan can >all initiate that prodding. > >A funny thing that I've seen happen across multiple brands of drive is where >the SMART historical data indicates that a drive was in very serious >trouble, as a monitored parameter would drop down to the lowest possible >(worst) value. Later on, for whatever reason, the drive recovered to perfect >health. Most recently, a 40GB WD hard disk in a Dimension 2400 showed this >behavior, with its "raw read error rate" dropping to 0. But at some point, >it recovered to perfect health. Why and how this could happen baffles me, >but I've seen several drives do it and go on to work fine for quite some >time. They're not drives I would trust, so they don't go into roles where a >drive failure would cause a real problem. > >SMART would probably work a lot better if there were more consistency in how >it was implemented, if drive makers were more honest and didn't mind their >drive looking bad if it meant an advance warning of disaster, and if all >drives gave their self test routines a higher priority. > >All of the above makes the assumption that SMART is enabled AND that someone >is listening. Most drives can disable their SMART diagnostics upon receiving >a command to do so. Some ASUS motherboard BIOSes (and others) do this! Of >course, even if the drive's SMART system is enabled, someone has to be >listening when it calls for help. Very few systems do that...most Dell >desktops do, as do the good old Compaq Deskpro EN systems. I've even seen a >few eMachines whose BIOS was watching for SMART problems, amazingly enough. >It's usually the BIOS that sounds the alarm, so a user only gets a warning >at power on time. Later SMART alerting would have to come from software that >knows how to find and interpret SMART data, as the BIOS can't do much after >the OS has taken control. > >I've seen SMART warnings save the day. If only that sort of thing happened >more often... > >William > Thanks William. You deserve it with this long "informative" post. I mean this in respect, not scarcism. Thank you.
From: RnR on 8 Feb 2010 19:48 On Mon, 08 Feb 2010 18:46:26 -0600, "RnR" <rnrtexas(a)gmail.com> wrote: >On Mon, 8 Feb 2010 18:34:21 -0600, "William R. Walsh" ><newsgroups1(a)idontwantjunqueemail.walshcomptech.com> wrote: > >>Hi! >> >>> Has SMART improved because from my past experience, SMART info was >>> wrong? >> >>Yes and no. >> >>Every drive maker has a different idea about how SMART should be >>implemented, how errors should be detected and what categories they should >>watch. And the thresholds all differ as well--which is to say that each >>manufacturer has a different idea of "how bad" things have to get before the >>drive is considered to be failing. >> >>Out of all of the drives I've seen, I'd say that Maxtor has the most >>informative SMART reporting. Yes, believe it or not. Their drives monitor a >>wide variety of parameters and the values do shift around while the drive >>operates. Second in line would be Seagate, whose drives don't show as many >>parameters but still seem to be pretty honest. Quantum drives fit somewhere >>in between, and Quantum was actually one of the first companies to implement >>SMART in a hard drive. >> >>Hitachi GST/IBM drives have a similar selection of monitored parameters, but >>I'll have to admit that I've never seen them vary a whole lot, even on very >>obviously sick drives. ExcelStor drives behave very similarly, which isn't >>too surprising considering that A) their drives are mostly Deskstar clones, >>B) HGST made some of their control boards and C) someone told me that HGST >>owns ExcelStor now. >> >>Western Digital (who makes otherwise excellent hard drives in my experience) >>has typically been very conservative with SMART data...with only a few >>monitored parameters and seemingly little variation, even on drives that >>were starting to have problems. >> >>Of course, SMART data is sometimes only collected by a drive when it is >>idle. In a modern operating system, between power management events (spin >>down) and disk activity in general, a drive may never get a chance to do its >>SMART self test routines because it either isn't idle for long enough, or >>it's spun down. (Other drives seem to observe what's going on while they are >>active, such as drives from Seagate, HGST and Maxtor.) So sometimes a drive >>needs a little prodding to update its SMART data and notice a problem. That >>kind of defeats the purpose. Tools such as SpinRite, HDAT2 and SpeedFan can >>all initiate that prodding. >> >>A funny thing that I've seen happen across multiple brands of drive is where >>the SMART historical data indicates that a drive was in very serious >>trouble, as a monitored parameter would drop down to the lowest possible >>(worst) value. Later on, for whatever reason, the drive recovered to perfect >>health. Most recently, a 40GB WD hard disk in a Dimension 2400 showed this >>behavior, with its "raw read error rate" dropping to 0. But at some point, >>it recovered to perfect health. Why and how this could happen baffles me, >>but I've seen several drives do it and go on to work fine for quite some >>time. They're not drives I would trust, so they don't go into roles where a >>drive failure would cause a real problem. >> >>SMART would probably work a lot better if there were more consistency in how >>it was implemented, if drive makers were more honest and didn't mind their >>drive looking bad if it meant an advance warning of disaster, and if all >>drives gave their self test routines a higher priority. >> >>All of the above makes the assumption that SMART is enabled AND that someone >>is listening. Most drives can disable their SMART diagnostics upon receiving >>a command to do so. Some ASUS motherboard BIOSes (and others) do this! Of >>course, even if the drive's SMART system is enabled, someone has to be >>listening when it calls for help. Very few systems do that...most Dell >>desktops do, as do the good old Compaq Deskpro EN systems. I've even seen a >>few eMachines whose BIOS was watching for SMART problems, amazingly enough. >>It's usually the BIOS that sounds the alarm, so a user only gets a warning >>at power on time. Later SMART alerting would have to come from software that >>knows how to find and interpret SMART data, as the BIOS can't do much after >>the OS has taken control. >> >>I've seen SMART warnings save the day. If only that sort of thing happened >>more often... >> >>William >> > > >Thanks William. You deserve it with this long "informative" post. >I mean this in respect, not scarcism. Thank you. Forgot to mention, yes I read every word of your post !!! thanks....
From: Sam on 8 Feb 2010 21:42 "Steve W." <csr684(a)NOTyahoo.com> wrote in message news:hkn22e$ueg$1(a)speranza.aioe.org... > Sam wrote: >> > > Using an 8200 to type this. > This one had a freezing problem a LONG time ago. (still in warranty) at > the time I ran through a lot of tests and determined it was a bad > motherboard. > Dell sent me an updated one and it has been fine since. > Although I have replaced the OEM video card and the hard drives and the > fan. Other than that it just keeps going. Not the fastest machine in the > world but for the web and E-Mail it''s fine. > > FYI the 8200 can handle up to 1 gig of memory. > Good to hear your 8200 is still working. I remember now it always had the lock up problem but Dell blew me off becuase I had partitioned the HDD. Eventually I changed it to dual boot Win200 and XP Pro. Still had the lock up problem then I changed the Win2000 partition to Linux and finally stopped using it becuase Linux wasn't ready for prime time. I'm not going to spend anymore money on it. s
From: Sam on 8 Feb 2010 21:46 "Ben Myers" <ben_myers(a)charter.net> wrote in message news:hkkqhf$oh$1(a)news.eternal-september.org... \ > > The Dimension 8200 system requires PC800-45 (or faster) RAMBUS memory. To > go to 2GB, you would need 4x512MB modules, which are pretty expensive > these days compared to DDR2 or even DDR. Before you invest in more > memory, I suggest that you take the time to figure out the cause of the > freezing up. There may be a hardware issue elsewhere, such as a hard > drive that is not in good health... Ben Myers I blew out the case and the fan in the power supply but it still locks up. I'm not spending more money on it.
From: Sam on 8 Feb 2010 21:48
"Daave" <daave(a)example.com> wrote in message news:w4-dneTQkf-7WvPWnZ2dnUVZ_rOdnZ2d(a)cavtel.net... > I would imagine something else is responsible for the freezes. Low RAM > just means certain tasks take significantly longer since the pagefile is > heavily used. > > If something else is responsible for the freezes, and if you can resolve > that problem, you might want to first determine if there is a need to > install more RAM. The only time this is necessary is if lots of paging > occurs. A quick way to determine if this is happening is to open Task > Manager (Ctrl+Alt+Del) and click the Performance tab. Then note the three > values under Commit Charge (K): in the lower left-hand corner: Total, > Limit, and Peak. > > The Total figure represents the amount of memory you are using at that > very moment. The Peak figure represents the highest amount of memory you > used since last bootup. If both these figures are below the value of > Physical Memory (K) Total, then you probably have plenty of RAM. > In case you want to explore this further, you may run Page File Monitor > for Windows XP: > > http://www.dougknox.com/xp/utils/xp_pagefilemon.htm I'll try your suiggestions. I decided not to spend anymore money on the PC. > > |