1

I'm trying to figure out what the read cache part of the following message refers to:

[sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA

Is it "IDE drive's read-lookahead feature (usually ON by default)" as treated by hdparm?

If it's not, how can this be turned off?

Mark Henderson
  • 68,823
  • 31
  • 180
  • 259
poige
  • 9,448
  • 2
  • 25
  • 52
  • Hi Poige. I suspect the -'ve votes are because what's in the title of your question isn't actually in the question. I'll tidy it up a bit for you. – Mark Henderson Dec 01 '11 at 07:55
  • @Mark Henderson♦, can you respond via moder's tool? – poige Dec 01 '11 at 10:33
  • New badge, Gold badge: You've earned the "Famous Question" badge (Question with 10,000 views) for "What is read cache in “[sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA”?"., earned on Server Fault – poige Jul 25 '19 at 04:49

1 Answers1

1

Well, I've found out proper tool to control this cache: sdparm.

It looks like as not having any relation to read-ahead cache. Also, seemingly rare SATA drive is able to unset this bit.

poige
  • 9,448
  • 2
  • 25
  • 52
  • 2
    What is your problem anyway? What do you want to achieve when meddling with this settings? – Sven Feb 01 '11 at 03:38
  • 2
    There's no problem at all, I just paid attention to this very dmesg's message. Also, I thought that turning read cache off in favor of write-cache could be a worth-doing thing if that would increase write cache accordingly (OS caches reading anyway). – poige Feb 01 '11 at 03:49
  • [Here](https://linux.goeszen.com/what-is-doesnt-support-dpo-or-fua.html) I learned that these are more obscure features of the ATAPI protocol most drivers won't support anyway. Don't know why the devs decided to put this debug message in. – isync Dec 04 '21 at 20:55