AFP share drops unexpectedly, but CIFS share is fine

[ScuseMe1]

Hi,

I don't know if this is a known problem with AFP shares, since I just started using them. But I have an AFP share to my music on an NV+ that's been dropping connection without warning. When I try to reconnect, I get this error dialog:



My CIFS share connects fine. But when I unmount the CIFS share and try to mount the AFP share, I still get this error. FrontView shows both protocols active on the share. I'm running RAIDiator 4.1.7, and as I said, this is the first time I'm using AFP shares with my iMac & MBP:





I'm running 802.11n to access the NV+, but my wifi doesn't drop, nor is anything else wifi related affected. Only the AFP share.

Any ideas? Is anyone else experiencing my problem?

[tstout]

I'm seeing this as well on a Pro Pioneer. Seems to be happening more since the latest firmware update. Appleshare just randomly stops working, but CIFS keeps going. Haven't figured out if there is a specific event triggering the failure or if it's just happening after a period of time.

For now, going into FrontView and disabling then re-enabling the Appleshare filing protocol seems to be the only fix I've come across.

[MarkF52]

Have you tried running it hard wired for a while, just to rule out wireless? I know you mentioned that you're wireless didn't drop but it can't hurt to try it just to eliminate some variables.

[ScuseMe1]

tstout wrote:

I'm seeing this as well on a Pro Pioneer. Seems to be happening more since the latest firmware update. Appleshare just randomly stops working, but CIFS keeps going. Haven't figured out if there is a specific event triggering the failure or if it's just happening after a period of time. For now, going into FrontView and disabling then re-enabling the Appleshare filing protocol seems to be the only fix I've come across.

Well, I'm sorry you're having problems, but I'm glad I'm not alone 🙂

For now, I'm just using the SMB/CIFS share, since my Sonos system needs that share to access my music on the NV+. Of course, I've had no problems with the CIFS share dropping or any other funky behavior. But, on my system, the AFP share is definitely faster, so I would love to use it. But I can't take the freezes.

[ScuseMe1]

MarkF52 wrote:
Have you tried running it hard wired for a while, just to rule out wireless? I know you mentioned that you're wireless didn't drop but it can't hurt to try it just to eliminate some variables.

I may be able to do that. I have a 100 ft roll of CAT5E which I can run from the gigabit router to my iMac. But I'd need to run the wires so that no one will trip over them. That's why wireless rules the roost here 🙂

But as I said, the AFP share becomes unresponsive. After I unmount it, I still can't log back into it; I get a connection error. The CIFS share - which is also mounted - is working fine. I have to go into FrontView and disable/re-enable AFP (as tstout said) to get the share working again. That doesn't sound like a wifi problem (since the CIFS share is still working with wifi), but anything's possible.

[tstout]

MarkF52 wrote:
Have you tried running it hard wired for a while, just to rule out wireless? I know you mentioned that you're wireless didn't drop but it can't hurt to try it just to eliminate some variables.

I'm running both an iMac and Mac mini wired already (laptop is wireless) so I don't think that's the issue on mine. Happens on all three at the same time.

Happened again last night when I got home from work, BTW -- everything appears normal; AFP share is still listed in the sidebar even, but get a failure when trying to connect to it (guest access is on for the share so it shouldn't be a password or session timeout issue).

Restarting AFP on the ReadyNAS immediately fixes the problem and luckily my wife rarely uses it. Gonna have to move the kid's movies to local storage or go through the CIFS share if it keeps up much longer, however.

I'm going to try disabling CIFS entirely for a while and see if that impacts the AFP stack (maybe some sort of protocol interference in the network stack -- I know, I'm reaching here, but there's not a lot to try.) If I figure it out, I'll let you know.

[ScuseMe1]

tstout wrote:
I'm going to try disabling CIFS entirely for a while and see if that impacts the AFP stack (maybe some sort of protocol interference in the network stack -- I know, I'm reaching here, but there's not a lot to try.) If I figure it out, I'll let you know.

For a couple of days, I only had the AFP protocol running (I wasn't using my Sonos system then), and I still got the hangs. That's what prompted me to reinstate CIFS and start using it. So good luck with disabling CIFS. If it's what I saw, it'll change nothing in the stability of AFP.

[mrwiggles]

I can say without hesitation that the most current ReadyNAS firmware (Intel at least) has a problem with the AFP services just dropping out periodically. I run into this ALL THE TIME. I attempt to mount an AFP volume using the FInder to connect to my Pro Business and all I get is NOTHING. CIFS works, but I don't use that and iTunes won't work with it. The only fix is stopping and restarting AFP services.

So, Yoh-Dah and Han, please acknowledge that you guys have an issue here and give us some idea of the plan for the fix.

-mw

[mrwiggles]

Another AFP drop out this morning. Just got a spinning beach ball and the same error message as the OP. This time it seemed to reconnect on its own, unlike past drop outs.

Between this and the iSCSI problems, I'm growing increasingly concerned about ReadyNAS Mac support.

-mw

[ScuseMe1]

mrwiggles wrote:
Another AFP drop out this morning. Just got a spinning beach ball and the same error message as the OP. This time it seemed to reconnect on its own, unlike past drop outs. Between this and the iSCSI problems, I'm growing increasingly concerned about ReadyNAS Mac support. -mw

Same here. I've temporarily given up on AFP until this is sorted out. My CIFS share has been rock solid.

[pilala]

- same here as well, ReadyNas Duo + Mac OS 10.6.5
- also using Sonos in combination with CIFS no problem,
- but AFP I have to manually disable and enable to have it appear in Finder again,
- as I am autoswitching off Readynas for the nite, this is most annoying,
- as it means for Timemachine to find the share the next day, I have to disable/enable the AFP service in Frontview every day!

one clue as to how this started: i upgraded the firmware on the Duo from "say a year old" to "most recent".

pretty please, with sugar on top, dear Lucaphiles, fix this .... please.

[osopolare]

I have the same issue with my ReadyNAS Pro Pioneer RAIDiator 4.2.15.

Each day we lose AFP and it has to be restarted. Time Machine also breaks.

[pilala]

So, so there seems to be an AFP issue then with the new firmware for OSX, as experienced by multiple users.

Who knows how to bring this under the attention of the the Readynas developers?

How long would it typically take to fix?

Thank you.

[Balders1]

I wouldn't think there's much the Netgear guys can do to address the AFP issues (hence their silence on why Mac support is so poor). I would assume they are waiting for a newer version of netatalk (we're currently on 2.1.3).

As it happens, 2.1.5 is out today. 🙂

All attempts by me to upgrade to this new version have failed (Linux isn't my strong point unfortunately) and I've given up after the 14th firmware re-install... Hopefully the Netgear chaps can test 2.1.5 and provide an update asap which 'may' help us out...

[silvein1]

I am having this issue now as well. Everything was fine until I put in a new hard drive and restarted the ReadyNAS to allow it to grow the volume.

Update: I SSH'ed in to the ReadyNAS and looked at netatalk's logs. It looks like it is shutting down on a signal 15, which is weird because I don't know why that would be issued, and by what.

Update 2: I wonder if the syslog restarts every day have anything to do with it. It seems to happen every day at the same time, which is the same time that netatalk stops. Maybe log rotation related?

Update 3: I reinstalled the last OS update and now I don't have this issue. Weird.

junction:/var/log# grep -hE 'Dec [0-9]{2} 06:25.*(shut|start).*' netatalk.log messages | sort | tail -n 15
Dec 20 06:25:08 junction syslogd 1.4.1#18: restart.
Dec 21 06:25:09 junction syslogd 1.4.1#18: restart.
Dec 22 06:25:09 junction syslogd 1.4.1#18: restart.
Dec 23 06:25:10 junction syslogd 1.4.1#18: restart.
Dec 24 06:25:08 junction syslogd 1.4.1#18: restart.
Dec 25 06:25:08.771068 afpd[19311] {main.c:104} (I:AFPDaemon): shutting down on signal 15
Dec 25 06:25:13 junction syslogd 1.4.1#18: restart.
Dec 26 06:25:02.566070 afpd[12691] {main.c:104} (I:AFPDaemon): shutting down on signal 15
Dec 26 06:25:12 junction syslogd 1.4.1#18: restart.
Dec 27 06:25:06.018825 afpd[32357] {main.c:104} (I:AFPDaemon): shutting down on signal 15
Dec 27 06:25:09 junction syslogd 1.4.1#18: restart.
Dec 28 06:25:05.280444 afpd[1981] {main.c:104} (I:AFPDaemon): shutting down on signal 15
Dec 28 06:25:09 junction syslogd 1.4.1#18: restart.
Dec 29 06:25:06.613803 afpd[24748] {main.c:104} (I:AFPDaemon): shutting down on signal 15
Dec 29 06:25:10 junction syslogd 1.4.1#18: restart.

[tino1]

I am having the same problem from two different NAS devices, NASDUO and NASNV+. I noticed after I installed the latest firmware 4.1.7. It's happened on two different macbooks (one is osx10.6.5 and the other is osx10.5). However CIFS works great, but you can't get time machine to work over CIFS, at least I don't think so. And we need time machine to work consistently.

NETGEAR please HELP!

[maximilianschei]

same here. but have also with cifs problem. sometimes both connections seems to drop.
just wanted to signal, that obvisiously many people having these problems. hope netgear will help soon. how to activate them?

best regards
max

[Green2]

I have the same problem!

I use two macs: iMac and an Macbook Pro. Both with OSX 10.6.5.
I have the Netgear DUO. It worked fine with AFP until the updated firmware came...

Has anyone issued this case to Netgear?

[osopolare]

I called Netgear support about the issue and the tech I spoke with seemed confident that he could fix the issue with a configuration change... after I bought a support contract.

Honestly I wasn't aware that software support was limited to 90 days. Oh well.

Maybe someone with an active support contract could call in and find out what the fix is and share it with us benighted and unsupported users.

[contempt1]

So is there a resolution to this? I NEVER experienced this problem on my NV+. But now my 3 week old Ultra 6 Plus just hit this bug for the first time tonight. I fixed it by shutting down AFP and restarting it but there's gotta be something the jedi's can do to fix this.

[silvein1]

I updated my previous post, but I guess its just not that visible.

I fixed this issue by reinstalling the firmware. Just go to update and tell it to update anyway if it is the same version.

[Green2]

I tried to re-innstall the firmware on my ReadyNAS DUO... That did not help... 😞

[knick620]

I am having exactly the same problems, just reinstalled firmware and watched the shares disappear as I tried to connect to them. I did have a relatively stable CIFS share, although 2 monitor avatars, a mac and pc, which made little sense. I am unable to create a timemachine backup as the afp share does not stay around long enought - any more ideas?

[dbfreeman_me_co]

I have experienced the same issue, but fortunately it happens pretty rarely (once every month or so). Also, happens since the last firmware upgrade. I haven't pulled the logs yet, but I did e-mail Netgear support asking for help with the issue and they were pretty useless. Perhaps someone from Netgear will be on this forum with some input. I've tested and confirmed that it is purely related to the AFP service dying on the ReadyNAS. Wired/wireless makes no difference. CIFS continues to work fine. This is annoying!

[pointpeninsula]

Has anyone found a reliable fix for this? I've had AFS and Timemachine running flawlessly on both a Macbook Pro and and iMac for months, and as of Saturday March 5th, I've been having these same problems. I'm really worried that my TimeMachine backups are gone, because I can't find them using SSH, and I keep all my working files on the NAS, as opposed to having them duplicated on both machines (and therefore backed up to both TimeMachine backups.)

Thanks in advance,

Tom