pveperf results

[gerdnl]

Are these normal results?

root@pve03:/var/lib/vz/dump# pveperf

CPU BOGOMIPS: 100793.76
REGEX/SECOND: 1018518
HD SIZE: 94.49 GB (/dev/mapper/pve-root)
BUFFERED READS: 54.46 MB/sec
AVERAGE SEEK TIME: 20.25 ms
FSYNCS/SECOND: 38.67
DNS EXT: 43.78 ms
DNS INT: 245.51 ms (myproxmoxdomain.com)

root@pve03:/var/lib/vz/dump#

somehow my backups are much slower lately, was hunting for issues

 

[tom]

Are these normal results?
...

Depends on your hardware and your configuration but yes, this is not fast. Provide all details about your hardware and storage setup.

 

[gerdnl]

supermicro server with 32gig mem
24x intel xeon(R) CPU E5-26220 2.10ghz 2 sockets
local storage to sata

4 nodes in my cluster from which 2 are identical, 1 is newer then the 2 and 1 is super new and running on ssd
the node i gave the info on is node 3 this is where the backups seems to stall a bit lately

 

[tom]

what is "local storage to sata"? if you want that we comment on this, you need to provide details.

- HDD or SSD specs
- used file-system

If you talk about the backup, you should also tell details about the backup storage.

 

[gerdnl]

root@pve03:~# parted /dev/sda print
Model: SMC SMC2208 (scsi)
Disk /dev/sda: 1998GB
Sector size (logical/physical): 512B/4096B
Partition Table: msdos
Disk Flags:

Number Start End Size Type File system Flags
1 1049kB 537MB 536MB primary ext3 boot
2 537MB 1998GB 1997GB primary lvm

Vendor: SMC
Product: SMC2208
Revision: 3.29
User Capacity: 1,997,998,653,440 bytes [1.99 TB]
Logical block size: 512 bytes
Logical Unit id: 0x600304800cf1b60019f82cbdc33785dc
Serial number: 00dc8537c3bd2cf81900b6f10c800403
Device type: disk
Local Time is: Thu Sep 7 13:25:31 2017 CEST
Device does not support SMART



The backup is stored on a QNAP NAS with uplinks to the network

 

[gerdnl]

the backup from last night wich i cancelled late during the day
https://dpaste.de/FKCW/raw

 

[gerdnl]

end part it didnt copy all
https://dpaste.de/6sN5/raw

 

[tom]

is this a raid controller? SMC2208. if yes, how did you configure you disk/raid?

looks like there is no writeback cache.

 

[gerdnl]

I think it is, raid1 i believe
not sure about the write back

 

[gerdnl]

0a:00.0 RAID bus controller [0104]: LSI Logic / Symbios Logic MegaRAID SAS 2208 [Thunderbolt] [1000:005b] (rev 05)

 

[gerdnl]

raid status
http://paste.debian.net/984978/

 

[tom]

looks like you have a hardware raid controller but no BBU (battery/flash) and therefore no write-back cache.

therefore you see the low fsyncs/seconds.