Is your X1000 unstable, and sometimes hangs on bootup?

I noticed my X1000 hardly ever booted fully, or locked up shortly after boot into WB when it was cold started. Sometimes it was just unstable in general. Hang or crash for no reason.

Solution: Reseat the memory module and all PCIe cards!

The memory was glued in my X1000, which means I never removed it for 10 years. So the contacts get corroded. Removing the memory and put in again, and the same with all PCI cards and SATA connectors made it much more stable, and not hang on cold boot.

Tip

@Deniil

The only problem I had was in 2020 : CR2032 battery (and when I tried to change it, the support broke, i had to replace it).

Changed and no problem at all since then.

@Deniil

Just remeber to nuplug the power, I bet its not RAM issue..
unless you want to spend a month hunting down some old DDR2 ram, with right specs.

CFE error about no ram, is a general error, used if it does find CDROM or HD.

@Deniil

I too have been frustrated by stability problems with my X1000 recently. Replacing the CR2032 battery (which had run down to a very low voltage) and reseating SATA cables seems to have solved the problem.

Before I checked the SATA cables, I had also reseated RAM, put it in different slots and tried different DIMMs, partly because of the misleading error messages given from CFE when there are drive errors, but I don't think the RAM was an issue in the end.

@Deniil

Funny, a number of years ago my X1/XE suddenly came down with a case of the black screen of death. Powered on but no sign on screen. On serial all I could see was a message about VGA then it stopped. Obviously suspecting my ATI card was at fault I check it over but it seemed fine. I then slowly pulled other things off like IDE cables and finally RAM. All the time checking serial. By the end I had it all stripped down and found out it could be stripped down to the bare minimum which was good as UBoot would still attempt boot and error out. After I removed the RAM and put it back it suddenly came back to life. Turned out the contacts had gone stale and simply disturbing the contact was enough to break it back into order. RAM was holding it back. I've only seen this on PPC machines.

Now, this is on the X1000. My X1000 has had various problems over the years. Here's a list:
* Failing to turn on or off.
* Stalling on boot logo after reset.
* CFE unable to detect HDD.
* CFE crashing during Kickstart load.
* Sudden SATA errors while system was in use.
* Sudden loss of SATA HDD after reset.
* Freeze on Workbench boot.
* USB EHCI driver broken and temporarily locking up system.

The power switch going faulty, failing to turn or off, was due to low battery. A bad power cable can also cause issues.

The stalling is an old CFE bug.

The inability to detect HDD, CFE crashing on Kickstart and other sudden SATA errors remain as unknowns. But, to stabilise this, I replaced my SATA cables and also installed SATA clips to secure them better over long periods. I also replaced my RAM which I bought new at the time. Should have bought spares. Since these proactive methods to stabilise system I haven't seen sudden SATA errors occur for a while.

The last two items. Unfortunately Update 2 brought me freezes on boot and Update 3 combined this with a broken EHCI driver. Apparently this only affects specific hardware configurations. I suspect a change in Exec or some other module has introduced a bug that has messed up interrupt signals or similar. That only shows up with particular hardware. I've confirmed in my case a screen mode change hangs on a vertical interrupt, which goes forever missing, and my Workbench hangs. Given it's interrupt related makes me wonder if the EHCI driver is also failing due to an interrupt issue.

I'll second removing RAM and reseatting the RAM sticks helped on my X1000.

I had X1000 unstable any process crashing soon after power on. I saw something about newlist or node assert failing displayed on debug info, so that made me to think problem with RAM. Both RAM modules had some dust in the middle of them. It looked a bit yellowish so I thought is there any component leaking. But luckily it was dust only. I cleaned them and put one module back. Now the machine is very stable again. Even gcc doesn't crash any more.

The Fractal case has filters in front of all holes and fans, but it gets some dust inside anyway. Cleaning the machine twice a year is maybe a good idea. Maybe a bed room is not a good place to have electronics, because changing clothes makes the most of dust.

That CFE's memory check command gives the same false values always no matter if the RAM works good or not. Some memory checker utility on OS4 is better to use, I guess.

Trying to find compatible DDR2 RAM might be a problem nowadays if the both modules would break. * knock knock on wood *

@LiveForIt

Quote:

Of course I unplug power
It wasn't the RAM itself, but the DIMM connectors. The slot pins. Microcorrosion on the contacts. Removing the DIMM module and putting it back solved the problem. Not replacing.

There was never any error message, or early boot failure, just general instability which got a lot worse shortly after a cold startup, which lead me to think it's probably a connector problem.

@ncafferkey

Good to know the battery also can cause instabilities. Never replaced that one. Typically run 24/7 so not much off time.

@Deniil

You should change the CMOS battery if you never changed it yet as once it gets a little low you'll notice boot up problems and glitches. You can check over at the Hyperion forum and you'll find lots of threads about this issue.