Kernel Error

vcr58

Noticed on the monitor this error message which was also in the system general log...

May 19 09:00:00	php	12611	[pfBlockerNG] Starting cron process.
May 19 09:00:26	php	12611	[pfBlockerNG] No changes to Firewall rules, skipping Filter Reload
May 19 09:06:18	kernel		MCA: Bank 2, Status 0x8400004040080151
May 19 09:06:18	kernel		MCA: Global Cap 0x0000000000000c07, Status 0x0000000000000000
May 19 09:06:18	kernel		MCA: Vendor "GenuineIntel", ID 0x706a1, APIC ID 2
May 19 09:06:18	kernel		MCA: CPU 1 COR (1) ICACHE L1 IRD error
May 19 09:06:18	kernel		MCA: Address 0x80db1530
May 19 09:21:00	sshguard	74510	Exiting on signal.

There were no performance issues as far as I know, everything seems normal.

Is this something I should be concerned about?

Thanks

stephenw10

@vcr58 said in Kernel Error:

Is this something I should be concerned about?

Yes. MCA errors like that are almost always a hardware issue.
Usually I would have expected it to panic and reboot after seeing that but I assume that did not happen?

Is this the first time you've seen an error like that?

Have you recently updated?

Steve

vcr58

@stephenw10 Correct, there was no reboot after this error. I only noticed it because I had a monitor attached, then I checked the system/general logs.

This is the first time I saw this error since replacing the SSD that was giving me trouble (see post). Nothing else has changed except I started using pfBlocker a couple days ago.

I have been running 22.01 released version for a couple months.

revengineer

I have seen this error a couple of times over the years on my FreeNAS/TrueNAS server. In each case, it was bad RAM. Run a few cycles of memtest86 to nail this down, then replace the bad module.

EDIT: It could also be CPU, but I would try memory test first before blaming the CPU. Running CPU stress test with Prime 95 or AIDA64 would be the next step if memory checks out.

stephenw10

Yeah, it's pretty much always hardware. Sometimes you might start seeing it after an upgrade for example which looks like a software issue but that's usually because some new driver is now hitting the hardware issue.

Steve

jimp

$ mcelog --no-dmi --ascii --file mce.log
mcelog: Family 6 Model 122 CPU: only decoding architectural errors
mcelog: Family 6 Model 122 CPU: only decoding architectural errors
Hardware event. This is not a software error.
CPU 1 BANK 2
ADDR 80db1530
MCG status:
STATUS 8400004040080151 MCGSTATUS 0
MCGCAP c07 APICID 2 SOCKETID 0
CPUID Vendor Intel Family 6 Model 122 Step 1

Given that it was an L1 cache error and that's the cache on the CPU, then it's almost certainly a CPU problem and not RAM. Might be overheating as well but that seems less likely.

If it is a board with a removable CPU you can try re-seating the CPU in the socket. If it has a removable heat sink you could also try removing that, cleaning and redoing the thermal paste/grease/tape/whatever.

vcr58

@jimp My CPU is soldered on a mini ITX MB but the heat sink may be removable. However I have never seen CPU temp above 40 deg C so I don't think its an issue.

I read somewhere in these forums that there was a BIOS setting that fixed a users errors. I found a "Turbo Mode" in BIOS that I disabled so maybe that will help. I haven't seen any more errors since my first post.