I'm receiving parity errors on a regular basis now. I don't recall ever seeing them in the past. I was seeing a lot of call traces due to having a custom docker net, I believe I have resolved those issues but the parity errors continue to show up during parity-checks. Any help/advice is appreciated, server seems to be running fine otherwise.

veyron-diagnostics-20190114-1257.zip

If there were no unclean shutdowns start by running memtest, bad RAM is the number one reason for unexpected sync errors.

3 minutes ago, BrandonG777 said:

I'm receiving parity errors on a regular basis now. I don't recall ever seeing them in the past. I was seeing a lot of call traces due to having a custom docker net, I believe I have resolved those issues but the parity errors continue to show up during parity-checks. Any help/advice is appreciated, server seems to be running fine otherwise.

veyron-diagnostics-20190114-1257.zip

Are you running your parity checks as correcting or non-correcting?     If you are using non-correcting then the errors will persist until you run a correcting check.    If you keep getting new errors then that is more concerning.

Edited January 14, 20197 yr by itimpi

 

Yeah, they keep showing up despite being corrected each week. I never recall seeing them before running into the call traces though, like something got corrupted somewhere. 

 

Also, just noticed this...

I'm thinking that's indicative of a failing drive <sigh> ...

2 minutes ago, BrandonG777 said:

I'm thinking that's indicative of a failing drive <sigh> ...

No, usually a bad SATA cable, though a single error is no case for concern, if it keeps increasing replace the cable.

1 minute ago, johnnie.black said:

No, usually a bad SATA cable, though a single error is no case for concern, if it keeps increasing replace the cable.

What should I do about the continuous parity errors though? Like I said I don't recall seeing these on a regular basis prior to my docker call trace problems.

1 minute ago, BrandonG777 said:

What should I do about the continuous parity errors though?

 

16 minutes ago, johnnie.black said:

If there were no unclean shutdowns start by running memtest, bad RAM is the number one reason for unexpected sync errors.

 

Be careful about running a correcting parity check on a schedule.  If a drive starts to fail/has read errors during a parity check (which is very possible since the hardware is under higher stress when checking parity) you could end up with a failed disk and invalid parity.

How many parity errors are you seeing?

6 minutes ago, WizADSL said:

How many parity errors are you seeing?

Between 4 and 22 but they've consistently come up since my docker call traces a couple months back. Doing an extended smart test now, and plan on doing a full memtest once I'm at home.

 

EDIT: @WizADSL I will take your advice on the parity correction scheduling.

Edited January 14, 20197 yr by BrandonG777

42 minutes ago, BrandonG777 said:

I will take your advice on the parity correction scheduling.

In addition to not correcting on a schedule, weekly would be considered excessive by most. Any reason for that frequency?

Just now, trurl said:

In addition to not correcting on a schedule, weekly would be considered excessive by most. Any reason for that frequency?

Honestly, I don't remember my reasoning for that schedule. What would you recommend, monthly?

5 minutes ago, BrandonG777 said:

What would you recommend, monthly?

That is the usual.

Okay I've had a couple months with no parity errors. Not sure what I did to resolve this but memtest came up clean. I did blow the dust out and ensured all the SATA connectors were secure. I also changed the schedule to start earlier in the night.

Sent from my Pixel 2 XL using Tapatalk