Time machine backup keeps losing connection

[chrisidle]

Hi

Im using FreeNAS 11.2 to backup a mac using time machine.

I have setup FreeNAS with a 2Tb pool for the AFP share and Time Machine recognises the FreeNAS disc correctly and starts a backup.

Unfortunately, the backup stop after about 5 minutes, 10G of data transfer. If i re-start the backup, the same think keeps happening.

When i backup to an external hard drive, the backup completes successfully so im assuming that the issue is something in my FreeNAS setup.

Does anyone have any suggestions as to whare to look or what to check?

Thanks

 

[anodos]

You can start by checking /var/log/afp.log on the FreeNAS server. Perhaps increase log level for AFP by adding the auxiliary parameter: log level=. Possible values are severe, error, warn, note, info, debug, debug6, debug7, debug8, debug9, maxdebug. Maybe try setting to "debug" if there's not much in the existing log.

 

[Mlovelace]

There appears to be a limit with the TCP reassembly queue length that is negatively affecting Apple clients. Try increasing the sysctl variable to a higher limit and try again. You can do this by adding the following to the 'Tunables' section of the 'System' tab.

Code:

Tunables

Variable: net.inet.tcp.reass.maxqueuelen
Value: 11586
Type:  sysctl

Edit: Here is a link to the bug report dealing with the issue.

 

[chrisidle]

You can start by checking /var/log/afp.log on the FreeNAS server. Perhaps increase log level for AFP by adding the auxiliary parameter: log level=. Possible values are severe, error, warn, note, info, debug, debug6, debug7, debug8, debug9, maxdebug. Maybe try setting to "debug" if there's not much in the existing log.

Hi, I have checked the logs and i get this every time i try to backup:

Feb 12 15:31:51.593198 afpd[11639] {auth.c:235} (note:AFPDaemon): Login by mackbook (AFP3.4)
Feb 12 15:48:39.830544 cnid_dbd[9329] {main.c:407} (info:CNID): Checkpointing BerkeleyDB for volume '/mnt/Time/.AppleDB'
Feb 12 16:11:43.864934 afpd[11639] {auth.c:804} (note:AFPDaemon): AFP logout bymackbook
Feb 12 16:11:43.865616 afpd[11639] {afp_dsi.c:108} (note:AFPDaemon): AFP statistics: 3806848.37 KB read, 1165468.75 KB written
Feb 12 16:11:43.865632 afpd[11639] {dircache.c:615} (info:AFPDaemon): dircache statistics: entries: 674, lookups: 30009, hits: 26937, misses: 703, added: 3043,removed: 2369, expunged: 2369, evicted: 0
Feb 12 16:11:43.865671 afpd[11639] {afp_dsi.c:584} (note:AFPDaemon): done
Feb 12 16:11:43.866443 afpd[2615] {main.c:151} (info:AFPDaemon): child[11639]: done

The amount of date read or written changes but the process is the same?

Thanks

 

[chrisidle]

There appears to be a limit with the TCP reassembly queue length that is negatively affecting Apple clients. Try increasing the sysctl variable to a higher limit and try again. You can do this by adding the following to the 'Tunables' section of the 'System' tab.

Code:

Tunables

Variable: net.inet.tcp.reass.maxqueuelen
Value: 11586
Type:  sysctl

Edit: Here is a link to the bug report dealing with the issue.

Thanks.

Unfortunately, i has not helped.

 

[anodos]

What does console say on the Mac? Does it report an error?

 

[chrisidle]

What does console say on the Mac? Does it report an error?

There are no errors shown on the Mac.

Time machine shows "starting Backup" then XXX bytes of XXX bytes backed up and a little later "Stopping Backup".

Each time this happens, it I "Start Backup" the process will repeat and approximately 1-2G of data is transferred.

This is the first backup to FreeNAS and at the moment there is about 320Gb to complete.

Thanks

 

[seanm]

There are no errors shown on the Mac.

Time machine shows "starting Backup" then XXX bytes of XXX bytes backed up and a little later "Stopping Backup".

Each time this happens, it I "Start Backup" the process will repeat and approximately 1-2G of data is transferred.

This is the first backup to FreeNAS and at the moment there is about 320Gb to complete.

Thanks

When anodos says 'console' he means for you to open /Applications/Utilities/Console.app. Once opened, type 'TimeMachine' (no space) in the search field at the top-right, then start a Time Machine backup. You should see various information logged. Hopefully containing a clue to your issue.

 

[chrisidle]

Hi

I have found the following errors whilst running time machine backup to my FreeNAS server.

2019-02-14 15:24:44.943936+0000 Error: (-43) SrcErr:YES Copying /Volumes/com.apple.TimeMachine.localsnapshots/Backups.backupdb/christopher’s MacBook Pro/2019-02-14-151816/Macintosh HD/usr/share/emacs/22.1/lisp to (null)

2019-02-14 16:57:42.022062+0000 Error: (-36) SrcErr:YES Copying /Volumes/com.apple.TimeMachine.localsnapshots/Backups.backupdb/christopher’s MacBook Pro/2019-02-14-164940/Macintosh HD/usr/local/mysql-8.0.13-macos10.14-x86_64/lib/mecab/dic/ipadic_sjis/sys.dic to (null)

If i backup to a USB drive, the backup complete correctly.

Any Ideas?

Thanks

 

[anodos]

Hi

I have found the following errors whilst running time machine backup to my FreeNAS server.

2019-02-14 15:24:44.943936+0000 Error: (-43) SrcErr:YES Copying /Volumes/com.apple.TimeMachine.localsnapshots/Backups.backupdb/christopher’s MacBook Pro/2019-02-14-151816/Macintosh HD/usr/share/emacs/22.1/lisp to (null)

2019-02-14 16:57:42.022062+0000 Error: (-36) SrcErr:YES Copying /Volumes/com.apple.TimeMachine.localsnapshots/Backups.backupdb/christopher’s MacBook Pro/2019-02-14-164940/Macintosh HD/usr/local/mysql-8.0.13-macos10.14-x86_64/lib/mecab/dic/ipadic_sjis/sys.dic to (null)

If i backup to a USB drive, the backup complete correctly.

Any Ideas?

Thanks

Looks like your share may have become unmounted during the backup. (sending the backup to (null) seems wrong). What OS-X version is this?

 

[chrisidle]

Looks like your share may have become unmounted during the backup. (sending the backup to (null) seems wrong). What OS-X version is this?

OX Mojave Version 10.14.2

Connection the the FreeNAS server is via hardwired ethernet.

Thanks