SOLVED TimeMachine not connecting

[SilverJS]

Hi all,

OK, so down to the last thing to get everything working the way I want it. Worst is - it used to run just perfect!!! Grr...

Anyhow : Followed docs to a T, but my Mac (Mac Mini, circa 2008) doesn't see the TimeMachine share half the time. The other half, it will see it, but tells me the share is unavailable right as backup commences. (In those instances, I see a quick 2-second logon for "studiomac", my Mac Mini username, in the console.)

If I stop AFP and restart, half the time, I'll get this kind of spam every second or so :

Apr 14 16:59:12 freenas-server kernel: <118>Apr 14 16:59:12 freenas-server afpd[10106]: main: no servers configured
Apr 14 16:59:13 freenas-server netatalk[10083]: Restarting 'afpd' (restarts: 9)
Apr 14 16:59:13 freenas-server afpd[10108]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:13 freenas-server afpd[10108]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:13 freenas-server afpd[10108]: main: no servers configured
Apr 14 16:59:13 freenas-server kernel: <118>Apr 14 16:59:13 freenas-server afpd[10108]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:13 freenas-server kernel: <118>Apr 14 16:59:13 freenas-server afpd[10108]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:13 freenas-server kernel: <118>Apr 14 16:59:13 freenas-server afpd[10108]: main: no servers configured
Apr 14 16:59:14 freenas-server netatalk[10083]: Restarting 'afpd' (restarts: 10)
Apr 14 16:59:14 freenas-server afpd[10109]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:14 freenas-server afpd[10109]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:14 freenas-server afpd[10109]: main: no servers configured
Apr 14 16:59:14 freenas-server kernel: <118>Apr 14 16:59:14 freenas-server afpd[10109]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:14 freenas-server kernel: <118>Apr 14 16:59:14 freenas-server afpd[10109]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:14 freenas-server kernel: <118>Apr 14 16:59:14 freenas-server afpd[10109]: main: no servers configured
Apr 14 16:59:15 freenas-server netatalk[10083]: Restarting 'afpd' (restarts: 11)
Apr 14 16:59:15 freenas-server afpd[10111]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:15 freenas-server afpd[10111]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:15 freenas-server afpd[10111]: main: no servers configured
Apr 14 16:59:15 freenas-server kernel: <118>Apr 14 16:59:15 freenas-server afpd[10111]: dsi_tcp_init: no suitable network config for TCP socket
Apr 14 16:59:15 freenas-server kernel: <118>Apr 14 16:59:15 freenas-server afpd[10111]: no suitable network address found, use "afp listen" or "afp interfaces"
Apr 14 16:59:15 freenas-server kernel: <118>Apr 14 16:59:15 freenas-server afpd[10111]: main: no servers configured

Ideas...?

Thanks!

EDIT : Forgot to say, but the Mini is connected via WiFi. It used to work fine though....!! Up until I re-made my pool...

 

[BigDave]

Up until I re-made my pool...

Before anyone can help, I'm sure you're going to have to give details about
what you mean by that exactly...

 

[SilverJS]

I just meant I recreated it with all of the individual folders as datasets, and copied the contents back from the backups box. This was in potential preparation for a third freenas computer, as I wrote in my other thread - which, I just saw, you also responded to! :). (Thanks for that.)

So, that's the reason. What's most frustrating is that everything worked perfectly before.

Actually, now that I think about it - it WAS working after the pool was rebuilt! What happened is, two nights ago, we had a huge storm and the power cut out. It was my ups's first test - passed! - but the Mac was in the middle of its first, big backup (133 of 260 gigs). Things haven't been right since after that.

 

[SilverJS]

Hmmm - interesting...I did a reboot with the Initial Install environment, let the Wizard do its thing with TimeMachine, and it worked flawlessly. Guess I'm now back to re-configuring that initial install. =)

 

[SilverJS]

Did someone change this thread to Solved without my knowing...? (I changed it back.) Because it certainly isn't - I was waiting until initial upload to complete to change thread status, because I didn't trust the connection. Turns out I was right - halfway through my upload, the AFP share went belly up again. Last attempt will be a full clean install from CD.....:(

EDIT : So far, my attempts with a clean install have been successful, but I'll wait and see. AFP just seems fragile in general...

 

[BigDave]

Did someone change this thread to Solved without my knowing...? (I changed it back.) Because it certainly isn't - I was waiting until initial upload to complete to change thread status, because I didn't trust the connection. Turns out I was right - halfway through my upload, the AFP share went belly up again. Last attempt will be a full clean install from CD.....:(

EDIT : So far, my attempts with a clean install have been successful, but I'll wait and see. AFP just seems fragile in general...

Storm might have effected something not FreeNAS? Was EVERYTHING in your system plugged into the UPS?

 

[SilverJS]

Both FreeNAS boxes and the network switch for them, that's all. Both boxes shutdown correctly, by the looks of it.

But the router wasn't, if that's what you're asking...everything else has been working fine since though.

 

[BigDave]

It sounded as though the backup failure after a period of time would be hardware?
I'm stumped! Sorry.

 

[rogerh]

Might be worth trying the original Mac backup with an ethernet connection, in case WiFi glitches may be causing problems. Is the WiFi signal good?

 

[SilverJS]

It sounded as though the backup failure after a period of time would be hardware?
I'm stumped! Sorry.

No worries at all, I'm always grateful for help.

To RogerH : Yeah, that's precisely what I was thinking. I'll probably go and buy a 50ft Cat5E cable tomorrow, to get that initial backup going. Like I've said though, it was actually working all fine before....too bad. But yes, the WiFi signal is actually quite strong throughout the house, the Mac Mini shows full bars.

 

[SilverJS]

Just a quick update on this since my past post :

My main box was really starting to act up weird after I wrote this. As in :

- Permissions were all screwed up. I started having this SID permissions thing that some people had a long time ago (and that I never had before). I couldn't add, delete, or change anything on my main Windows share;
- Replication completed, but with a slew of errors (like, hundreds). All of them were something about key read uudecode failing or some such. Ironically enough, all of the data was there, accessible, working just fine;
- Started getting these weird UPS errors about data being stale - on both Master and Slave boxes;
- I tried rebooting, and it sometimes would get stuck in this weird loop - just didn't trust it at all. Decided it was time for the clean install.

Downloaded the latest ISO (as of last night - FreeNAS-9.3-STABLE-201504100216), put it on a CD and installed it. Re-imported pool, all is good. I'm now using Replication as a bit of a stress test. My main and backups boxes are identical in terms of storage config (6 X 3TB, RaidZ2). I have about 3.7 TB of data, so one box can hold twice my data quite comfortably. On my backups box, I have two folders : one, where I manually dragged everything on the main box via Windows, and the other, as Replication target. I had started dragging everything over manually since yesterday when I started thinking something might be corrupt somewhere, and I started not even trusting replication. Turns out that was a good move. =)

So anyway - last night, after the clean install, I deleted the (error-stricken...?) Replication dataset on the backups box, and started a new replication. It's just wrapping that up now as I type. (EDIT : Completed, no issues.) But back on topic, I also set up the TimeMachine share last night as part of my initial config, and it's uploading just fine now, over WiFi. I know I said I'd get a network cable, but I just wanted to test my current setup to see if I could trust it. (It worked fine the first time, and I expect the same from it this time.) All errors above are now blissfully absent.

We'll see, but this looks like it wasn't a problem with FreeNAS or anything - quite likely on my side.

 

[Robert Trevellyan]

You might consider replacing those Kingston USB sticks. They have a bad reputation in this forum and a corrupt boot pool could cause all kinds of strange behavior. I believe the fact that a clean install fixed your problems lends weight to this hypothesis.

 

[SilverJS]

I thought of that too. I haven't really researched that - any recommended ones?

 

[Bidule0hm]

The SanDisk Cruzer Fit is well known to be cheap ($10 for the 16 GB version) but to be good quality and very small ;)

 

[SilverJS]

Just to update and cap this off: all is well. Backup completed,all other backups occurring on schedule. So, clearly a boot pool issue.