Hi, In the last 3 days, I have been testing the other kernels. The steps which I took: 1. Boot to kernel-4.15.0-136. 2. Install kernel X, where X = 136,137,139,140: sudo apt-get install linux-image-4.15.0-X-generic linux-headers-4.15.0-X-generic linux-modules-4.15.0-X-generic linux-modules-extra-4.15.0-X-generic 3. Boot to kernel X. 4. Manually run: sudo depmod 5. Reboot. 4. Run the server's normal workload and see how it was performing. I haven't seen disk performance issues with those kernels. But there is one difference, which I don't know how to explain. Before I raised this bug, we had the same disk performance issues on kernel-4.15.0-140. I collected the list of kernel modules running on that kernel at that time: Module Size Used by joydev 24576 0 input_leds 16384 0 hid_generic 16384 0 usbhid 49152 0 hid 122880 2 usbhid,hid_generic ip6_tables 28672 0 iptable_filter 16384 0 ip_tables 28672 1 iptable_filter x_tables 40960 3 iptable_filter,ip6_tables,ip_tables tcp_diag 16384 0 udp_diag 16384 0 inet_diag 24576 2 tcp_diag,udp_diag cpuid 16384 0 btrfs 1155072 0 zstd_compress 163840 1 btrfs xor 24576 1 btrfs raid6_pq 114688 1 btrfs ufs 77824 0 qnx4 16384 0 minix 32768 0 ntfs 102400 0 msdos 20480 0 jfs 188416 0 xfs 1204224 0 libcrc32c 16384 1 xfs intel_rapl 20480 0 sb_edac 24576 0 x86_pkg_temp_thermal 16384 0 intel_powerclamp 16384 0 coretemp 16384 0 kvm_intel 217088 0 kvm 614400 1 kvm_intel ipmi_ssif 32768 0 irqbypass 16384 1 kvm crct10dif_pclmul 16384 0 crc32_pclmul 16384 0 ghash_clmulni_intel 16384 0 pcbc 16384 0 aesni_intel 188416 0 aes_x86_64 20480 1 aesni_intel cdc_ether 16384 0 usbnet 45056 1 cdc_ether mii 16384 1 usbnet crypto_simd 16384 1 aesni_intel glue_helper 16384 1 aesni_intel cryptd 24576 3 crypto_simd,ghash_clmulni_intel,aesni_intel intel_cstate 20480 0 intel_rapl_perf 16384 0 ipmi_si 57344 0 ipmi_devintf 20480 0 ipmi_msghandler 53248 3 ipmi_devintf,ipmi_si,ipmi_ssif shpchp 36864 0 mei_me 40960 0 mei 94208 1 mei_me lpc_ich 24576 0 acpi_power_meter 20480 0 acpi_pad 180224 0 mac_hid 16384 0 bonding 163840 0 lp 20480 0 parport 49152 1 lp autofs4 40960 2 mgag200 45056 1 i2c_algo_bit 16384 1 mgag200 ttm 106496 1 mgag200 drm_kms_helper 172032 1 mgag200 syscopyarea 16384 1 drm_kms_helper sysfillrect 16384 1 drm_kms_helper tg3 167936 0 sysimgblt 16384 1 drm_kms_helper fb_sys_fops 16384 1 drm_kms_helper ptp 20480 1 tg3 drm 401408 4 drm_kms_helper,mgag200,ttm pps_core 20480 1 ptp megaraid_sas 143360 12 mxm_wmi 16384 0 wmi 24576 1 mxm_wmi Now, after the testing we are running on kernel-4.15.0-140 (again) and this time I don't see the issue. But the list of kernel module is different: lsmod Module Size Used by intel_rapl 20480 0 sb_edac 24576 0 x86_pkg_temp_thermal 16384 0 intel_powerclamp 16384 0 coretemp 16384 0 kvm_intel 217088 0 ipmi_ssif 32768 0 kvm 614400 1 kvm_intel irqbypass 16384 1 kvm crct10dif_pclmul 16384 0 crc32_pclmul 16384 0 ghash_clmulni_intel 16384 0 pcbc 16384 0 aesni_intel 188416 0 aes_x86_64 20480 1 aesni_intel crypto_simd 16384 1 aesni_intel glue_helper 16384 1 aesni_intel cryptd 24576 3 crypto_simd,ghash_clmulni_intel,aesni_intel intel_cstate 20480 0 cdc_ether 16384 0 usbnet 45056 1 cdc_ether input_leds 16384 0 joydev 24576 0 mii 16384 1 usbnet intel_rapl_perf 16384 0 ipmi_si 57344 0 ipmi_devintf 20480 0 ipmi_msghandler 53248 3 ipmi_devintf,ipmi_si,ipmi_ssif mei_me 40960 0 mei 94208 1 mei_me lpc_ich 24576 0 shpchp 36864 0 mac_hid 16384 0 acpi_pad 180224 0 acpi_power_meter 20480 0 bonding 163840 0 lp 20480 0 parport 49152 1 lp autofs4 40960 2 mgag200 45056 1 i2c_algo_bit 16384 1 mgag200 ttm 106496 1 mgag200 drm_kms_helper 172032 1 mgag200 syscopyarea 16384 1 drm_kms_helper sysfillrect 16384 1 drm_kms_helper sysimgblt 16384 1 drm_kms_helper fb_sys_fops 16384 1 drm_kms_helper tg3 167936 0 hid_generic 16384 0 ptp 20480 1 tg3 usbhid 49152 0 drm 401408 4 drm_kms_helper,mgag200,ttm pps_core 20480 1 ptp hid 122880 2 usbhid,hid_generic megaraid_sas 143360 12 mxm_wmi 16384 0 wmi 24576 1 mxm_wmi Perhaps those additional kernel modules which were loaded in the first time were affecting disk IOs?