Project

General

Profile

Bug #49576

Apple OSX backup very slow, eventually seems to stop and not to proceed

Added by Davide Violante 12 months ago. Updated 10 months ago.

Status:
Closed
Priority:
No priority
Assignee:
Andrew Walker
Category:
Services
Target version:
Seen in:
Severity:
New
Reason for Closing:
Reason for Blocked:
Need additional information from Author
Needs QA:
Yes
Needs Doc:
Yes
Needs Merging:
Yes
Needs Automation:
No
Support Suite Ticket:
n/a
Hardware Configuration:
ChangeLog Required:
No

Description

I have a Mac OSX High Sierra, i use Time Machine to backup .
I can easily add the Time Machine backup share using a local user, backup start to write to the share at 20Mbits over gigabyte LAN instead all windows backup are writing at 80Mbits and the speed remain costant. Instead Apple time Mchine backups starts to degrade to the point that the Mbits are 0 and no progress.

However no backup failures time Machine shows no progress just hanging in the for hours without progress.

I have a system with Dual core CPU on Intel Board and 16GB memory


Related issues

Related to FreeNAS - Bug #43558: Relax the TCP reassembly queue length limit to improve performanceDone

History

#1 Updated by Davide Violante 12 months ago

  • File debug-HOOFDDORPNAS-20181003144302.txz added
  • Private changed from No to Yes

#2 Updated by Dru Lavigne 12 months ago

  • Assignee changed from Release Council to Andrew Walker

#3 Updated by Ryan Moeller 12 months ago

  • Related to Bug #43558: Relax the TCP reassembly queue length limit to improve performance added

#4 Updated by Andrew Walker 12 months ago

Let's try two additional diagnostic steps:
1) Bump the value of net.inet.tcp.reass.maxqueuelen to 1437 by typing the following in the CLI and re-trying the backup:

sysctl net.inet.tcp.reass.maxqueuelen=1437

If that fails to resolve the issue, then:

2) Try increase netatalk's log level to "debug"

log level = debug

Once you have completed (2), please generate a new debug and attach to this ticket.

#5 Updated by Davide Violante 11 months ago

Andrew Walker wrote:

Let's try two additional diagnostic steps:
1) Bump the value of net.inet.tcp.reass.maxqueuelen to 1437 by typing the following in the CLI and re-trying the backup:
[...]
If that fails to resolve the issue, then:

2) Try increase netatalk's log level to "debug"
[...]

Once you have completed (2), please generate a new debug and attach to this ticket.

Thanks I am gona try this, however I can tell you the following.

The TimeMachine backups first initial full backup is 194GB it took in the end about 4 hours to backup. It seems that at 27 GB stops proceeding inrality din't sever stop, this time I let it continuing and after 30 mins of apparent idle, the backup continued. so 4 hours including that 30 mins of apparent idle

It seems that the speed on average was 15Mbits max, the Cifs backups are good, I reach 80 sometime 100 mbits on a gigabit network with cisco switches.

I let you know some results

#6 Updated by Andrew Walker 11 months ago

Hi Davide, do you have an update on how the testing went?

#7 Updated by Dru Lavigne 11 months ago

  • Status changed from Unscreened to Blocked
  • Reason for Blocked set to Need additional information from Author

#8 Updated by Dru Lavigne 10 months ago

  • File deleted (debug-HOOFDDORPNAS-20181003144302.txz)

#9 Updated by Dru Lavigne 10 months ago

  • Status changed from Blocked to Closed
  • Target version changed from Backlog to N/A
  • Private changed from Yes to No

Davide: please leave a comment on this ticket about your results if you get a chance to retest.

Also available in: Atom PDF