time machine very slow with mountain lion

I think I figured it out!!

My estimate was at 600 days (yes, almost two years).

I "excluded" Google Drive. I'm now at two hours!

This may be true for other networked drives. Just go into Time Machine Preferences --> Options and select all your network drives to exclude (including shared computers).

Please let me know if this also works for you.

Seems like so many solutions! try this

It is nothing to do with wifi speeds as I am running on a USB disk, upgraded from SL to ML recently but Time Machine had had other problems before then so hadn't used for a while.

I wanted to set up a new TM disk - 918GB back up. When I ran TM it took an hour to get going and then told me...53 days! Guess what I did? NOTHING

I left it running while I read all 14 pages of this forum. Then left it overnight.

18 hours later it was all done. All 918 GB of it. I think Apple have the same issues as Microsoft had with their 'estimated time' in Windows Explorer. I watched it go from 10 days at around 8am to 3 days by 9am and it kept going down and down from there.

TRY IT!

Hello dcjn

sorry but we went deeper in the analysis than just looking at extimated time.

When having the problem I left backup well above overnight...

For sure extimate time (not anymore available in ML) is less than linear because count on file number and not size of backup, so it is completely misleading. Also count that speed of backup in MB is much higher for large file than smaller one because of directory structure writing process so correct gauging/extimation duration is very difficult for MacOS as well as Windows.

As reported in previous post, for instance I went tracing the Network and seeing that all pending IO were completed and then network was almost silent (except some AFP control messages) for approx 3min and 20 seconds before a burst which size was variable (even few KB for small file) and the behavious stay consisten for 6 hours before I stopped the test.

I had impression that TCP stack was blocking communications, but I'm not so sure now, I'm more toward deadlocks between filesystems activities like Spotlight indexing or file permission and TM process itself.

regards, Tiziano

@jtc65 Same experience here. Apple Support isn't aware of Apple Marketing.

Apple has purposely inhibited Time Machine on network drives by throttling the bandwidth making it virtually useless (vs 'breakng'). They've done this to make it a "new" "feature" in Mountain Lion Server.

I guess they needed to give Mountain Lion Server more purpose by marketing it as your LAN's solution for all things backup related. What better way then by breaking Time Machine on regular OS X? <grin>

http://iansutherland.ca/2012/08/15/slow-time-machine-network-backups-in-mountain -lion/

Curiously, Time Machine works fine on shared Linux drives (if you setup a Netatalk, an Open Source version of Apple File System) on a Linux box.

http://pwntr.com/2012/03/03/easy-mac-os-x-lion-10-7-time-machine-backup-using-an -ubuntu-linux-server-11-10-12-04-lts-and-up/

Hello Scarrol

just to avoid someone enthusiasm

My NAS runs Ubuntu Server 12.04.1 LTS + Netatalk (2.2.1-1) since May with no TM problem till I've migrated to 10.7.5. NAS has latest kernel and is up-to-date to latest patches

Now, by doing Repair disk permission on all disks before _ upgrading to ML I've resolved the problem but I'm crossing finger. My son's Mac still on 10.7.5 still has the problem.

Backup is essentially unusable also on USB attached disk (sorry no firewire available)

Regards, Tiziano

Just adding another voice to the noise. Staring down a 157-day ETA on a time machine backup after days of tweaking & trying every trick in the book. Have allowed baks to run up to three days to assure no bad-reporting artifact, nope - the reported TM increment matches that apparent in DU at all times. MacBookPro on ethernet (WLAN shares issue); setup is local peer-to-peer network Mac+Win, LinkSys WRT300N & DLink switch at the core, all Cat6, Time Capsules in bridge mode. with wireless "on". Running OSX 10.7.5, prior to that no issues. Nothing new/obvious in console report nor in tmdiagnose.

What an unbelievable waste of time. Long history of giving up on Macs in response to head-in-sand issue avoidance (MBPro's & MBAir's) by Apple is accumulating to the breaking point. Tired of responding by buying more & complicating life. Immediate plan is use CrashPlan to local eSata GRaidMini, bak data/apps only. All this fiddling with terminal, console, arcane command-line stuff... the whole pitch for the Mac was "never have to open the hood, just drive".

My solution. It worked for me.

1: exclude from backups Google Drive, Dropbox and other frequently changed folders.

2: full disable spotlight like this

$ sudo su -

(enter password here)

# launchctl unload -w /System/Library/LaunchDaemons/com.apple.metadata.mds.plist

# mdutil -a -i off

Hi sorry, not worked for me.

i'm simply trying to back up, nothign complex going on, no complex set up. MBPro, ML recent update installed, and it won't back up. takes ages to prepare, only have 7.28Gb to back up, it gets to 2.24 and then doesn't progress. doesn't fail either, but doesn't progress beyond this. adn i can't leave it for days, as i need access to the machine.

i may try to leave it to back up for a day or so tonight though.

Well if there are 14 paegs and internet is full of this same error, google it you will be surprised, why Apple is not fixing it properly. It all worked fine with me till latest OS Lion update now I am doing 15 kb per sec up to 2 yrs of back up ..... Come on can no one fix this ?

Apple moderators please wake up !

One possible solution:

I an on a wired network and my time machine drive (a USB Drobo) is connected to a MacBookPro which I use as a backup server.

I have been experiencing extremely slow TM backups on my wired computers (weeks, months, years etc.), but normal speed backups on my wirelss computers.

I went to each wired computer and switched the network settings from DHCP to manual, each time entering the DHCP address assigned to the computer in the manual setting. The backups sped up instantly. Went from 150k per second to 20MB per second (as viewed in Activity Monitor) instantly as each computer was changed.

For me it's a problem that goes beyond Time Machine. It's a problem that affects the speed of communication between a Mac with ML and any external HD. I'm using Crash Plan+ to back up to my external HD using FireWire and the speed is just 4Mbps!!!!

I have been having all these same problems and my Time Machine HD is one of the four INTERNAL HDs in my 10.7.5 Mac Pro.

I have given up on TM for now and rely on Carbon Copy Cloner.

Maybe Apple will take this debacle seriously some day.

Bill

Just adding my [dejected] voice to this list. Am still on 10.7.5 Running on a early 2009 17" MBP - cannot seem to resolve the problem.

Cet

😟

Well it has now run for over more than 16 hours and got 6 GB of teh 182, so only 18days to go ........

Wake up guys this is serious, did you got lost on the way to the office on IOS6 Maps ?

Bill thanks for your post, I'll stop investing nights and weekends into investigating my network setup now based on your report.

I remain astounded at the Apple silence policy in this and similarly-unacknowledged prior issues (how many of us have MBP's purchased in resignation to replace cooked-thermal-paste prior model?). There's so much goodwill from Apple afficionados that no marketing damage would be sustained by courteous handling of the issue, viz.:
1. We acknowledge there is a problem.

2. (Describe the phenomenon and to the extent possible document who is affected).

3. We are determined to expeditiously resolve the problem.

4. In the meantime you should _________________

5. We promise to post timely updates at _____________ until resolved.

... which would provide me and I suspect most others with sufficient confidence to continue buying Mac.

Is it a stock-shock sensitivity? Just corporate culture or a divisional quirk? Who knows. What I do know is: issue disclosure is now the prerogative of users, all one has to do is broadcast frustrations factually on Twitter. What Apple is doing is failing to engage a priceless opportunity to manage issue impact/contagion.