Re: ReadyNAS Ultra 2 with OS 6.4.2 periodically drops network connection

tcmalone · ‎2016-03-17

Following up on an issue raised by several people at the very end (pages 127-8) of an old/closed topic:

"OS6 now works on x86 Legacy WARNING: NO NTGR SUPPORT!"

I am having the same problem as others where the Readynas is dropping off the network, and can't be found with RAIDar 6 or by web-interface at a static IP address. It seems to occur every 1-2 weeks and requires a power-off-reset to get it back on the network. I am running Firmware 6.4.2 and no installed apps. Services include: SMB, AFP, ReadyDLNA, UPnP, HTTP, HTTPS, and SSH.

Once the ReadyNAS is rebooted, it seems to work fine for a few weeks - and then dissapears!

mdgm-ntgr · ‎2016-03-17

Can you update to 6.5.0 Beta and see if you still have the same issue?

tcmalone · ‎2016-03-17

Ok, just updated to ReadyNASOS 6.5.0-T322 (Beta 1). Thanks.

Now, we'll see in a week or two if it drops off the network again... 😉

By the way, one of my use conditions is that I only use it for routine PC backups every few weeks - so it sometimes goes for days or weeks WITHOUT any client network access. I don't know if that is part of the issue.

tcmalone · ‎2016-04-18

After my upgrade to ReadyNASOS 6.5.0-T322 (Beta 1), and having verified it was working properly for a week - I left on a trip for two weeks and came home to find that again I can't "see" the ReadyNAS on my network. All other network devices, PCs, network printer, WiFi access points - all are fine. But using both Raidar6 and Win7 'Windows Explorer' I can't locate the NAS, nor does a browser pointed to the NAS IP address find anything.

There is no visible indication that the NAS is not working, i.e. the "Power Button" is solid 'blue' and the two "Drive" LEDs are solid 'green'. However, there is no activity on the "ACT" LED at all. My only recourse at this point is to manually power-down the NAS and restart.

Should I try upgrading to a new version again?? Thanks.

mdgm-ntgr · ‎2016-04-19

Can you send in a fresh set of logs (see the Sending Logs link in my sig)?

Are you using NIC bonding/teaming ?

6.5.0 Beta 2 does have some additional fixes e.g. there is a fix for an issue with the anti-virus service so that would be worth a try.

tcmalone · ‎2016-04-25

mdgm, I have just sent the logs by email. And will now upgrade to 6.5.0 RC1 (not Beta 2), and continue testing. Thanks.

mdgm-ntgr · ‎2016-04-26

Let us know how your testing on 6.5.0-RC1 goes.

If you need to do another unplanned reboot download the logs shortly after that and send them in.

tcmalone · ‎2016-05-18

I've just had to restart (hard power down) the NAS due to the same condition - not visible on network or RAIDar 6 after two+ weeks of inactivity. After the restart I downloaded the logs and just emailed them to you.

A few weeks ago (after your last post), I was going to use/test RC1, but then in the beta area I saw a release with no extension and installed that. I thought it was the final 6.5.0, but I guess not as my ReadyNAS is now asking me to upgrade from 6.5.0 to 6.5.0. I suppose you can check the log files I just sent to see what firmware that really was.

In any case I've just let the ReadyNAS upgrade the firmware on it's own, and will now let it run for another few weeks and see if it can "stay visible" on the newwork.

By the way, my client's ReadyNAS - same exact machine with 6.4.x also exhibits this same behavior, dropping from network and disapearing from any network app (like Sonos) after a few weeks of inactivity!

Thanks,

Todd

mdgm-ntgr · ‎2016-05-19

Which apps are you running?

You are using disks with a known firmware bug. Unfortunately SeaGate hasn't released a firmware update for ST2000DL003-9VT166 disks that shipped with CC32 firmware, only for disks that shipped with CC3C firmware. CC3D firmware (or later) works fine, but only disks with CC3C firmware can be updated to CC3D.

tcmalone · ‎2016-05-21

Earlier, I had not been running any "Apps", and still there was this issue. A month ago I downloaded DDclient, but that has since been removed. I mentioned, my client is also having the same issues with his ReadyNAS Ultra 2 running OS 6.4.x, and that unit is not running any Apps either.

I've read quite of few posts now about the issues with the Seagate ST2000DL003-9VT166 disks shipped with CC32 firware - but it seems to affect drives by dropping from the RAID array. Are you saying you think this might affect BOTH drives at the same time and cause my NAS to hang completely? Since these drives are obviously out-of-warranty, do you have any other suggestions for upgrading/replacing them?

I will check with my client to see what drives/firmware he is using to help compare issues.

StephenB · ‎2016-05-22

@tcmalone wrote:
Since these drives are obviously out-of-warranty, do you have any other suggestions for upgrading/replacing them?

I recommend NAS-purposed drives (WDC Red or Seagate VN series).

tcmalone · ‎2016-06-07

After another few weeks running but with no interaction (i.e. not accessing/reading/writing to the NAS) I tried to find the NAS with Raidar with no luck. So obviously I'm still having the same problems as before, even with the final release of 6.5.0.

I'm still very curious about my earlier post/question on 5/21 to 'mdgm' about whether you think this "network connection" problem is related to my using the Seagate ST2000DL003-9VT166 disks with CC32 firware. I still don't understand how having a drive drop from the RAID array would cause the OS to stop responding to the network. Again, are you saying there is 'evidence' or a good possibility that this drive issue could cause my NAS to hang completely?

Thank you.

Todd

StephenB · ‎2016-06-08

@tcmalone wrote:

I'm still very curious about my earlier post/question on 5/21 to 'mdgm' about whether you think this "network connection" problem is related to my using the Seagate ST2000DL003-9VT166 disks with CC32 firware. I still don't understand how having a drive drop from the RAID array would cause the OS to stop responding to the network. Again, are you saying there is 'evidence' or a good possibility that this drive issue could cause my NAS to hang completely?

In general, drive issues can cause a lot of performance problems and sometimes boot failures.

But I think you'll need @mdgm-ntgr to respond with the evidence he saw. Perhaps PM him directly. It might take a little while before he can get to it.

tcmalone · ‎2016-08-13

Following up again after several months. As mentioned before, this loss of network connection after a period of inactivity happens to some of my client NASs as well. And their ReadyNAS Ultra 2 is not using the same "problematic" Seagate ST2000DL003-9VT166 disks with CC32 firware. These are the drives in my client's Ultra 2 running OS 6.5.0-T338:

Seagate ST2000DM001-1CH164

So, what do you make of this? I don't think this is a drive issue. Again, the problem arrises after a week or two of network inactivity. The NAS is continously powered and on the network, but without any periodic or scheduled activities. I will update his firmware to 6.5.1 today, and wait another two weeks, and check the status again.

Thanks,
Todd

sotrack · ‎2016-08-13

I faced such issue with my NAS more than year ago . Problem was router Linksys E4200. I have replaced one and forgot the problem.

Wrong router configuration can be RCA. What is your NAS static address? And what is your DHCP range of addresses? Is static address out of this range?

tcmalone · ‎2016-08-13

My router is a TP-Link Archer C7. It's local IP is 192.168.1.1, DHCP range 192.168.1.100-199; and NAS static IP set to 192.168.1.20. I have other static devices set between 192.168.1.5-30

However, my client who also experiences this problem, has a completely different router (AT&T Uverse modem/router) and has the NAS set to use DHCP.

I never experienced this "network connection loss" due to inactivity before upgrading my Ultra 2 to OS 6.x. Previously with the older RAIDiator 4.2.x and earlier, I could ignore (not use) my NAS for weeks/months and still come back to see it on the network. Now, I lose the connection after a few weeks of not using it.

sotrack · ‎2016-08-13

OK. I use static IP address assigned to NAS MAC addres on router side currently. NAS is not aware his address is static, it use DHCP.

With Linksys was used NAS static IP and "connection loss" happened sometime. Static address was out of DHCP range.

mdgm-ntgr · ‎2016-08-14

Can you send in a fresh set of logs (see the Sending Logs link in my sig)?

Note that here is the option to put RAIDiator 4.2.x back on your Ultra if you want: https://community.netgear.com/t5/Community-Add-ons/Downgrade-to-RAIDiator-x86-from-ReadyNAS-OS-notsu...

You would have to backup your data first as the disks would need to be wiped.

mdgm-ntgr · ‎2016-09-18

Do your routers have gigabit (1000Mb/s) ethernet ports?

Both NAS units are only connected at 100Mb/s. It would be best if the NAS was using a gigabit ethernet connection.

Also ReadyNAS OS 6.5.2 is now available!

tcmalone · ‎2016-09-18

Possibly, as listed in a previous post my router is a TP-Link Archer C7, and so thank you for the suggestion to improve performance... But, hopefully you're not implying that could have something to do with my overall issue of network connectivity??

mdgm-ntgr · ‎2016-09-18

Users have run into issues in the past using a 100Mbit connection with the Ultra 2, so it would definitely be worth checking if the problem is still seen if the Ultra 2 has a gigabit connection.

tcmalone · ‎2017-01-03

mdgm: Ok, after another month or so of testing the two Ultra 2 units (both running 6.6.0), the one connected to a gigabit Ethernet connection does NOT "drop off" the network, while the other Ultra 2 using a 10/100 hub still does periodically (roughly every 2 weeks) disappear from the network (and needs to be rebooted). I will now swap locations/connections of the two units and see if this pattern continues to be true.

However, if this is "known" issue (i.e. Ultra 2 devices having issues on 10/100 networks), why isn't Netgear doing anything to resolve this? I hardly think it is fair or reasonable to expect all consumer users to provide a gigabit connection, when not all routers are gigabit to begin with, and also cheap 10/100 hubs are still commonly in use.

Honestly, is the Ultra 2 so far out of production (end-of-life/discontinued) that there simply will NEVER be any further support for it? Thanks!

mdgm-ntgr · ‎2017-01-03

We released our first OS6 models over 3 and a half years ago now.

RAIDiator-x86 4.2.x is EOL. If it gets any further updates they would be primarily security fixes.

Running OS6 on the Ultra 2 is not officially supported. Using a gigabit connection works fine and more and more routers and switches support gigabit these days. To get the most out of your NAS you do need to use a gigabit connection. A 100Mbit connection would be a bottleneck.

Rootax · ‎2017-01-03

Do you really use a hub ? Have you tried at least with a switch ?

StephenB · ‎2017-01-04

@Rootax wrote:

Do you really use a hub ?

That's a good point. Hubs broadcast all the traffic they receive to all ports they are connected to, and when there is a speed mismatch they definitely are subject to congestion - which will hurt performance and provoke dropped connections. A lot of the old 10/100 hubs couldn't even keep up with fast ethernet at full wire speed (even though the interfaces could negotiate 100 mbits).

Even if it is a fast ethernet switch, the best solution is to replace it with a gigabit switch (spending the $30 or so that costs). That solves the problem and improves network performance.