This is a discussion on Ultra 60 question within the Sun Solaris Hardware forums, part of the Solaris Operating System category; --> All, I have an ultra60 that came with 4x64MB in Bank 1 and 4x32MB in bank 3 (rendering bank ...
| |||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| All, I have an ultra60 that came with 4x64MB in Bank 1 and 4x32MB in bank 3 (rendering bank 3 offline since one stick was not being recognized). I got hte machine jumpstarted 2 days ago, then last night decided to replace the bad stick and add another bank of 32MB SIMMs since I had them lying around. I noticed that the new (3rd) bank of memory was physically hot to the touch, so I pulled it out. Now I get the following; the banner shows 384MB of memory, but when I try to boot the machine off disk, it just hangs; when I try to boot it off the network, I get the following: {0} ok boot net Timeout waiting for ARP/RARP packet 4200 And it just stops at 4200; doesn't count any higher. I also noticed that with diag-level set to max, when the ECC/etc tests are being run, the INFO lines report 0MB in bank 1 and 128MB in bank 3, but then after POST, the machine seems to see all 384MB. Any ideas what, if anything, I fried? I'm going to try some more diagnostics (moving memory to different banks, etc) when I get home from work, but any pointers would be very helpful. (And Rich, this means NNTP is working again [for now], so I don't have to bug you so much Thanks, -- Coy Hile hile@cse.psu.edu |
| |||
| It seems like one of memory stick keep make problem. You may need to turn on "diag-switch?" as "true" to see more detail information. If you can available any console terminal as like a Hyper term or Wyse it would be the best way to See all Post Diagnostic messages and find out which one keep makes trouble. Solar "Coy Hile" <hile@cse.psu.edu> wrote in message news:cl5q63$2fb$1@neuromancer.cse.psu.edu... > All, > > I have an ultra60 that came with 4x64MB in Bank 1 and 4x32MB in bank 3 > (rendering bank 3 offline since one stick was not being recognized). I > got hte machine jumpstarted 2 days ago, then last night decided to > replace the bad stick and add another bank of 32MB SIMMs since I had > them lying around. > > I noticed that the new (3rd) bank of memory was physically hot to the > touch, so I pulled it out. > > Now I get the following; the banner shows 384MB of memory, but when I > try to boot the machine off disk, it just hangs; when I try to boot it > off the network, I get the following: > > {0} ok boot net > Timeout waiting for ARP/RARP packet > 4200 > > And it just stops at 4200; doesn't count any higher. > > I also noticed that with diag-level set to max, when the ECC/etc tests > are being run, the INFO lines report 0MB in bank 1 and 128MB in bank 3, > but then after POST, the machine seems to see all 384MB. > > Any ideas what, if anything, I fried? I'm going to try some more > diagnostics (moving memory to different banks, etc) when I get home from > work, but any pointers would be very helpful. > > (And Rich, this means NNTP is working again [for now], so I don't have > to bug you so much > > Thanks, > > -- > Coy Hile > hile@cse.psu.edu |
| |||
| In article <cl5q63$2fb$1@neuromancer.cse.psu.edu>, Coy Hile <hile@cse.psu.edu> writes: > All, > > I have an ultra60 that came with 4x64MB in Bank 1 and 4x32MB in bank 3 > (rendering bank 3 offline since one stick was not being recognized). I > got hte machine jumpstarted 2 days ago, then last night decided to > replace the bad stick and add another bank of 32MB SIMMs since I had > them lying around. > > I noticed that the new (3rd) bank of memory was physically hot to the > touch, so I pulled it out. > > Now I get the following; the banner shows 384MB of memory, but when I > try to boot the machine off disk, it just hangs; when I try to boot it > off the network, I get the following: > > {0} ok boot net > Timeout waiting for ARP/RARP packet > 4200 > > And it just stops at 4200; doesn't count any higher. > > I also noticed that with diag-level set to max, when the ECC/etc tests > are being run, the INFO lines report 0MB in bank 1 and 128MB in bank 3, > but then after POST, the machine seems to see all 384MB. > > Any ideas what, if anything, I fried? I'm going to try some more > diagnostics (moving memory to different banks, etc) when I get home from > work, but any pointers would be very helpful. > > (And Rich, this means NNTP is working again [for now], so I don't have > to bug you so much > > Thanks, > Hi, have never seen that non-existing memory is reported. You certainly have fried the memory controller. -- Michael Tosch IT Specialist HP Managed Services Technology Solutions Group Hewlett-Packard GmbH Phone: +49 2407 575 313 Mail: michael.tosch:hp.com |
| ||||
| Just an update, I pulled the 4x64 out (that had been somewhat flaky when I first got the machine). Now the machine boots and such just fine with 4x32 in bank 3. It's entirely possible that I fried bank 1, but I think it was the 64MB sticks since putting them into bank 3 had POST fail saying it coudln't find any memory. Thanks for all your suggestions. -- Coy Hile hile@cse.psu.edu |