Anatoly Pugachev <matorola@xxxxxxxxx> writes: > just checked /home filesystem, there's no errors... And yes, there's > sometimes (sporadically) I see kernel messages like those in logs: Interesting. In case it helps, I have accurate timestamps for the three events we've seen so far: 2021-04-22 20:16:43 MSK 2021-05-05 20:25:52 MSK 2021-05-07 03:31:39 MSK The last one correlates with what you showed here: > May 07 03:31:39 gcc202 kernel: dm-0: writeback error on inode > 2148294407, offset 0, sector 159239256 > May 07 03:31:39 gcc202 kernel: sunvdc: vdc_tx_trigger() failure, err=-11 > May 07 03:31:39 gcc202 kernel: blk_update_request: I/O error, dev > vdiskc, sector 157618896 op 0x1:(WRITE) flags 0x4800 phys_seg 16 prio > class 0 Also, while this might be a mirage, I have the distinct impression that the problem is load-sensitive. Yesterday I could not make the failure happen just by looping the test case that had shown it previously. But after realizing I was the only one using the machine, I tried starting some non-conflicting tests in additional ssh sessions, and then it happened pretty quickly with the extra load. regards, tom lane
