UDMA CRC ERROR COUNT

[FraxTech]

A while back one of my drives that is in a disk shelf threw a UDMA CRC ERROR COUNT of 1.  It has stayed at 1 for a couple of months, so I'm not worried about it, but now every time I log into the server I have error messages about it.  I was wondering if there is any way to just acknowledge the error and have UNRaid stop throwing it unless it changes?

 

I've attached a copy of my diagnostics in case that is of any help.  Thanks, everyone.

epcot-diagnostics-20210616-2257.zip

[remotevisitor]

See last sentence in the UDMA CRC Errors section from the user manual.  A link to the user manual can be found at the bottom right of the Unraid UI.

Edited June 17, 2021 by remotevisitor

[FraxTech]

9 minutes ago, remotevisitor said:

See last sentence in the UDMA CRC Errors section from the user manual.  A link to the user manual can be found at the bottom right of the Unraid UI.

Thanks remotevisitor.  I'm sorry, I guess I should have stated, I've done that a number of times, but it keeps turning back to orange and then showing the error on the right side of the OS when I log in.  Is there any way to make sure that once I've acknowledged the error, it remains that way unless it increases?

 

 

[itimpi]

7 hours ago, FraxTech said:

Thanks remotevisitor.  I'm sorry, I guess I should have stated, I've done that a number of times, but it keeps turning back to orange and then showing the error on the right side of the OS when I log in.  Is there any way to make sure that once I've acknowledged the error, it remains that way unless it increases?

 

 

If you have acknowledged the error then it should not prompt you again unless the value changes.  
 

 This suggests there is something else going on but I am not sure what.    Maybe someone else will have a suggestion?

[Jackal]

Hi all.

 

As this is my 1st message in this forum, I am short of introducing myself directly with a question at hand !!!  I am a new user of UNRAID and have built a small NAS for storage purposes. I converted my old PC, an i7-4770, and added 4 new HDDs (2 x Seagate IronWolf & 2x Toshiba N300) along with a couple of old HDDs and an SSD that I had laying around. The HDDs are attached to an LSI card 9211-8i which I converted to HBA with the newest FirmWare (this forum provided excellent help and I appreciate that !!! )

mpt2sas_cm0: LSISAS2008: FWVersion(20.00.07.00), ChipRevision(0x03), BiosVersion(07.39.02.00)

 

However, yesterday I got informed that 2 of the brand new drives, 1 IronWolf & 1 Toshiba, have "UDMA CRC ERRORs". One has 5 while the other one has 72. I checked the logs and I did not see anything else suspicious. UNRAID does not reveal any Read/Write Errors. I do not like this at all. Especially for these 2 HDDs that are brand new. I am kindly asking for your ideas what things to check in order to minimize these CRC errors in the future. I understand it is not critical but it is not a good thing either.

 

Below, are some more information.

1. The LSI card was bought 2nd hand from e-bay. The seller had very good reputation and insured me that the card was bought locally from a store.
2. The cables that were included were brand new in a sealed box
3. During the firmware update with the LSI official FirmWare, all went smoothly
   1. I even put a fan on top of the card as seen here  (https://www.thingiverse.com/thing:4171229)
4. My PSU is 550W (which I do not suspect that this could be the problem)
   1. I would be more suspicious that all the HDDs are in a common cable and not in two different lines like the PSUs that can be found in Workstation PCs
5. I pre-cleared all the HDDs without any problems before adding them in the array.

 

Your ideas are more than welcome!!! Thanks.

 

monkeyisland-diagnostics-20210624-1907.zip

Edited June 24, 2021 by Jackal

[itimpi]

CRC errors are rarely actual drive problems.    They are connection problems and are typically related to the SATA or power cabling.   Normally they will just cause a retry of a read or write and as long as that succeeds no error is reported (although the retries can downgrade performance while they are taking place).  It is also worth noting that the CRC count can never be reset to zero - you only know that there is no issue remaining if the counts remain stable.

[Jackal]

And now I am afraid that 1 of the 2 parity disks I have went offline Why

This part was in Yellow !!! 

Jun 24 19:45:48 MonkeyIsland kernel: sd 7:0:3:0: [sde] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=0x00

 

And this part was in RED !!!

Jun 24 20:30:37 MonkeyIsland kernel: md: disk29 read error, sector=3907077192
Jun 24 20:30:37 MonkeyIsland kernel: md: disk29 write error, sector=3907077192
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 read error, sector=24
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 read error, sector=1501708064
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 read error, sector=1501708072
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 read error, sector=1501708080
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 read error, sector=1501708088
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 write error, sector=24
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 write error, sector=1501708064
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 write error, sector=1501708072
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 write error, sector=1501708080
Jun 24 20:31:07 MonkeyIsland kernel: md: disk29 write error, sector=1501708088

 

What is going on? I just created a folder ;-) !!! Nothing More !!!

 

What shall I do now ? ? ?