Any activity on HDD is become veeeery slow

[Sidius]

Hello
Running proxmox ve about one year on 12 x 12th Gen Intel(R) Core(TM) i7-12700 (1 Socket). This is not high loaded VE (one VM, two LXC). I have one nvme for VE itself, one SSD for high speed VM/LXC and one HDD (Seagate Barracuda st2000dm008-2fr102 2tb) for VM/LXC which do not need high speed disk access.All storages is lvm-thin. Several days before i need to run another VM, and i see that installation on HDD is veeeeeery slow (after waiting about 30 minutes and see that process is on 35% i kill VM, remove vm disk from hdd and put it on ssd and all was quick). I'm start investigate this problem. Create test LXC (debian 11), and start apt-get update, apt-get upgrade. It still veeeeery slow ( lasting about 14 minutes). When it running i looking on htop disk io - it's growing to 100%+ and LA of VE about 3-4. I'm checking hdd write speed in lxc with command dd if=/dev/zero of=test.file bs=64M count=16 oflag=dsync
it shows 1073741824 bytes (1.1 GB, 1.0 GiB) copied, 13.2918 s, 80.8 MB/s

I have another proxmox server with HDD. Trying do the same on it - apt-get update && apt-get upgrade is lasting only one minute not 14 minutes on first proxmox.
the same dd check on second promox server is show 1073741824 bytes (1.1 GB, 1.0 GiB) copied, 5.82312 s, 184 MB/s

Ok hdd is, looking on dd check, twice speed but not 14x speed when do any operation (((


When i do all of this on SSD - all is ok, very quick.

How to resolve/investigate this problem without buying new HDD?

 

[Pifouney]

Hey,

Check your SMART attrributes for your disk.
On your VE Host, check (gui) system -> System log

 

[Sidius]

Mainly messages like this

Sep 30 04:52:13 px smartd[853]: Device: /dev/sda [SAT], SMART Usage Attribute: 190 Airflow_Temperature_Cel changed from 68 to 69
Sep 30 04:52:13 px smartd[853]: Device: /dev/sda [SAT], SMART Usage Attribute: 194 Temperature_Celsius changed from 32 to 31

Several messages like this

Sep 30 04:22:12 px smartd[853]: Device: /dev/sda [SAT], SMART Prefailure Attribute: 1 Raw_Read_Error_Rate changed from 82 to 83
Sep 30 04:22:12 px smartd[853]: Device: /dev/sda [SAT], SMART Usage Attribute: 195 Hardware_ECC_Recovered changed from 82 to 83

In the SMART attributes - nothing special.

 

[carles89]

In the SMART attributes - nothing special.

It seems your disk is starting to fail, since logs show Raw_Read_Error_Rate and Hardware_ECC_Recovered values are increasing. I would definitely replace the disk.

apt-get update && apt-get upgrade

Don't use apt-get upgrade. Use always apt dist-upgrade to avoid dependency problems.

 

[Sidius]

This smart attributes is not a problem, on another VE - several the same errors and all is working fine. HDD is new (about one year), VE is on UPS and do not crash shutdown. Any other suggestions?

 

[leesteken]

Seagate Barracuda st2000dm008 is an SMR drive and might need more idle time to reorganize the drive as it is used more? People have complained here in the past that SMR drives can become very slow. Not sure what to do about it, except letting it run idle for a long time or replace it with a CMR drive.

 

[Sidius]

Thanks. I undestood. Trying to replace it with CMR drive.