Dear Chris, Am 16.01.20 um 02:57 schrieb Christoph Anton Mitterer: > On Wed, 2020-01-15 at 23:04 +0100, Oliver Freyermuth wrote: >> I have recently upgraded to 5.4.11 from a 5.3 kernel and now also hit >> the dreaded read time tree block corruption > > Is there some known corruption bug in 5.4? I don't think so / don't know, but the tree checker became more sensitive as compared to 5.3, so I dreaded this message after I saw the density of such reports from other users growing on the list in the past months. My only overlap with an earlier report seems to be that I had a situation of high memory pressure in the past weeks on the machine. Two other machines I run with very similar setups and significantly longer runtimes have not shown any issues. The affected node has the lowest runtime of the three, "but it ran fine with Kernel 5.4 for a while" and now suddenly detected the corruption in old data extens quickly after bootup. I don't find any issues with memtext86+. So I'll keep an eye on the machine. At this point, anything else would be guesswork. Cheers, Oliver
