MCE Hardware Error

[Friction9045]

I just looked at fix common problems and it told me about a machine check event being logged. I took a look through my logs and saw this:

Sep 21 11:18:23 Tower kernel: mce: [Hardware Error]: Machine check events logged
Sep 21 11:18:23 Tower kernel: [Hardware Error]: Corrected error, no action required.
Sep 21 11:18:23 Tower kernel: [Hardware Error]: CPU:1 (19:21:2) MC17_STATUS[-|CE|-|-|-|-|-|-|-]: 0x8000000100a08163
Sep 21 11:18:23 Tower kernel: [Hardware Error]: IPID: 0x0000000000000000
Sep 21 11:18:23 Tower kernel: [Hardware Error]: Bank 17 is reserved.
Sep 21 11:18:23 Tower kernel: [Hardware Error]: cache level: L3/GEN, tx: INSN

There's a also this that happened much earlier:

Aug 30 07:26:43 Tower kernel: mce: [Hardware Error]: Machine check events logged
Aug 30 07:26:43 Tower kernel: [Hardware Error]: Corrected error, no action required.
Aug 30 07:26:43 Tower kernel: [Hardware Error]: CPU:1 (19:21:2) MC21_STATUS[-|CE|MiscV|AddrV|-|-|-|Poison|-]: 0x8d480824548d4808
Aug 30 07:26:43 Tower kernel: [Hardware Error]: Error Addr: 0x0000000000000000
Aug 30 07:26:43 Tower kernel: [Hardware Error]: IPID: 0x0000000000000000
Aug 30 07:26:43 Tower kernel: [Hardware Error]: Bank 21 is reserved.
Aug 30 07:26:43 Tower kernel: [Hardware Error]: cache level: RESV, tx: GEN

Full log zip is tower-diagnostics-20240922-0215.zip

Is this anything to be concerned about?

[JorgeB]

If it keeps happening it may indicate an issue with the CPU cache memory.

[Friction9045]

50 minutes ago, JorgeB said:

If it keeps happening it may indicate an issue with the CPU cache memory.

I'll keep an eye on it then. If it keeps happening, does it mean I have to buy a new processor?

[JorgeB]

Can't say for sure the CPU is the problem, but it looks to me like the most likely culprit.

 

The errors are being corrected so far, since the cache memory uses ECC, but it's still not a good sign, and an uncorrectable error may crash the server.