Networking report shows period of VERY high activity

[cts]

I am seeing something weird and am not sure where to look. I have a FreeNAS Mini XL with 4x4tb raid10 disks - supporting NFS/iScsi (for now).

When i am moving large files (4gb -> 40gb) to the NAS via either NFS or iScsi i see good throughput and then there are periods of High network activity (maybe like retries???) and during those periods there is NO disk activity being reported.

I have attached a screen clipping of the network and disk activity.

Any thoughts?

 

[depasseg]

You can paste a screenshot directly into the post. Most security conscious folks are't going to open an unknown PDF.

 

[cts]

I will try to post the screen shot directly:

 

[depasseg]

My guess is that RAM is buffering the writes, which is why you don't see disk activity. And possibly why your throughput increases. Sync writes will need to be written to the pool, which tends to slow things down. If you throw some sync=disabled workloads into the mix (like possible your iscsi), then zfs will use RAM.

 

[Mirfster]

How much RAM you got? Do you have a good SLOG? *Sorry, don't know what the FreeNAS Mini comes with and too lazy to look it up...

I vote "no" for having sync=disabled, but to each their own. ;)

 

[cts]

the FreeNAS Mini XL comes with 32GB of ram. How do i check SLOG?

 

[Mirfster]

Notice mine shows a drive under "logs":

Code:

zpool status
  pool: ASC-ISCSI-V01
state: ONLINE
  scan: none requested
config:

  NAME  STATE  READ WRITE CKSUM
  ASC-ISCSI-V01  ONLINE  0  0  0
  mirror-0  ONLINE  0  0  0
  gptid/3c425ade-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  gptid/3cea0c4f-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  mirror-1  ONLINE  0  0  0
  gptid/3d9ded5e-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  gptid/3e4aec5a-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  mirror-2  ONLINE  0  0  0
  gptid/3ef3d0b4-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  gptid/3f999ec8-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  mirror-3  ONLINE  0  0  0
  gptid/4044d272-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  gptid/40eaae5b-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  mirror-4  ONLINE  0  0  0
  gptid/41939289-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  gptid/42376123-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  logs
  gptid/a96a9141-6bbb-11e6-90bb-000c293acb47  ONLINE  0  0  0
  spares
  gptid/9da42339-6bbb-11e6-90bb-000c293acb47  AVAIL
  gptid/9e4846e4-6bbb-11e6-90bb-000c293acb47  AVAIL

errors: No known data errors

That is my Intel DC S3710 200GB SSD which is being used as a SLOG

 

[cts]

interesting - when i enter zpool status i get:

Code:

[root@freenas ~]# zpool status                                                 
  pool: freenas-boot                                                           
state: ONLINE                                                                 
  scan: scrub repaired 0 in 0h0m with 0 errors on Thu Sep  8 03:45:17 2016     
config:                                                                         
                                                                               
        NAME        STATE     READ WRITE CKSUM                                 
        freenas-boot  ONLINE       0     0     0                               
          ada0p2    ONLINE       0     0     0                                 
                                                                               
errors: No known data errors                                                   
                                                                               
  pool: vg01                                                                   
state: ONLINE                                                                 
  scan: resilvered 21.9M in 0h0m with 0 errors on Wed Sep  7 07:53:56 2016     
config:                                                                         
                                                                               
        NAME                                            STATE     READ WRITE CKS
UM                                                                             
        vg01                                            ONLINE       0     0   
0                                                                             
          mirror-0                                      ONLINE       0     0   
0                                                                             
            gptid/c0425ec1-74f1-11e6-bca4-d05099c0ff2f  ONLINE       0     0   
0                                                                             
            gptid/608ea009-747d-11e6-9a01-d05099c0ff2f  ONLINE       0     0   
0                                                                             
          mirror-1                                      ONLINE       0     0   
0                                                                             
            gptid/618451e4-747d-11e6-9a01-d05099c0ff2f  ONLINE       0     0   
0                                                                             
            gptid/6276a1d7-747d-11e6-9a01-d05099c0ff2f  ONLINE       0     0   
0                                                                             
                                                                               
errors: No known data errors   

 

[Mirfster]

Then you don't have a SLOG. Not saying you *have* to have one, but I do since I run iSCSI with "sync=always" and would rather leverage a fast SSD. But tossing more RAM at it is preferable as well.

Some info for you:
https://forums.freenas.org/index.php?threads/some-insights-into-slog-zil-with-zfs-on-freenas.13633/
https://forums.freenas.org/index.ph...res-more-resources-for-the-same-result.28178/