Question NAS Shared Folder suddenly becomes inaccessible after some time ?

[Yossaris]

Hello everyone,

I couldn't find a thread detailing my exact problem, so I wanted to start my own in hopes of finding a solution.

As written in the title: on my company's Synology NAS Cluster (DS1817 - DSM 7.2.1-69057 Update 6) shared folders suddenly become inaccessible after some time. That period can range from days to hours and I think it depends on the amount of traffic in the subnet in that time. Interestingly enough, the folders can still be accessed from a different subnet that is not at the same physical location as the NAS, while a different subnet at the same location cannot access the folders. After doing a switchover between the passive and the active host, the access is restored.

The relevant parts of the network look something like this:

Subnet 1 - Users from Site 1 where the NAS is physically located
Subnet 2 - Where the NAS' IP-Address is located
Subnet 3 - Users from Site 2, who can still access the NAS, when users from Subnet 1 cannot

This problem is new and probably either started after an unexpected shutdown of the NAS (Blackout with a faulty UPS) or after the DSM Update, but imo it sounds more like a network issue, that I cannot pinpoint.

Does anyone have any experience with this kind of phenomenon? Any suggestion and help would be appreciated.

 

[Ralston18]

What error message(s) are presented when access fails?

Perhaps the unexpected shutdown corrupted files and/or did some physical damage to the NAS.

Did anyone start unplugging and replugging network connections? Swapping ports on the router or switches, etc..?

Another thought is that some of the configuration settings went back to defaults - especially if someone attempted to reset the NAS. Or other network devices. Or trying fixes that involved changing IP addresses and./or subnet masks. All too easy to have some typo wreak havoc.

Check all device IP addresseses (Static and Dynamic) and the corresponding sub-net masks.

I suggest sketching out a full network map showing all devices - not just what is deemed as the relevant partrs.

On the sketch include device names, macs, IP addresses (DHCP or Static) and subnet masks.

Seeing the proverbial "big picture" may help discover what has gone astray.

 

[Yossaris]

Thank you for the quick reply
The only error message that was relayed to me was: "The path is not available.
Cannot access X:\Path\to\folder. The file is currently not available on this computer." (Translated from German)
When put into google, the only solution popping up is that the offline cache is corrupted. But it seems unlikely for that to be the case if it is on every computer in the subnet, or so I assume.

I already went through all the switches and routers in all the subnets looking for some kind of mismatch, but there doesn't appear to be one. I can also access the webUI from everywhere, even when the shared folders are inaccessible.

I thought physical damage or corrupted files to be unlikely, since it works perfectly fine from a different subnet.

For a bigger picture:
Site 1 contains the following Subnets: 1 (for old devices that have not yet been moved to new subnets), 30 (NAS and Servers), 41&46 (User access via Ethernet and WiFi respectively), 42 (Test subnet I setup to check if the folders are still accessible from there)
Site 2 contains basically the same just with different "names" for the subnets (50,61&66)
The 2 Sites are connected by a bridge in a different subnet (155).

All the routing is done by D-Link 3130 Switches, while users are plugged into 1510 and 1210 Switches.

Site 1 users are routed from 41 or 46 to 30 and then gain access to the folders stored in the NAS.

Site 2 users are routed from 61 or 66 to 155 and then to 30.

 

[bill001g]

If it functions from any location it is not likely a issue with the NAS. I will assume the NAS setup is very simple and for example it is only on a single subnet so it sends all communications for other subnets to the router/default gateway.

Your best option is likely to run ping and tracert from various locations. If your router is smart enough you can many times tell it to ping with a source address on different subnets.

Key here is to try to find what is different from the ones that work and the ones the do not. I would also test to other devices on the same subnets. It would be extremely strange if it is only the NAS.

 

[Yossaris]

I will try those approaches as soon as the problem arises again, thank you.
I can't test too extensively every time since the shared folders are needed in the day-to-day business and need to be available as soon as possible, which makes troubleshooting this kind of annoying.

 

[Yossaris]

Update:
I had the opportunity to wait for a while before doing the switchover and noticed some more things:
While the folders remained accessible for a few minutes in the other subnet, that access was also lost after a while.
I could not access any folders on that particular cluster. Servers in the same subnet as the NAS seem to be accessible all the time.
I guess that points to the NAS being the issue after all. Is there anything I can check to make sure no settings or files got corrupted? The only indicator I can find is the health-status of the cluster, which says Healthy all the time.

 

[bill001g]

If you lose access from devices in the same subnet then you suspect the NAS itself. It would be strange that all devices did not lose it at the same time but I guess it depends on what is going wrong. You could rule out a network issue by leaving a ping command run. Just find a number of machines on different subnets and leave a constant ping run to the NAS ip address. This can just be in a background cmd windows so it will not disrupt the use of the machine for other purposes. You then fairly quickly when the problem happens check these machines. In this case since the problem does not fix itself you have a bit more time to check different machines.

If these work then you start to suspect the NAS. I would not think it was in the file system itself I would more be suspect of some kind of session table where it keeps a list of open session and for some reason it does not allow new files to be opened. Hard to say you are quickly getting into how the OS for a particular NAS system works. I would check the firmware levels and see if there are any reported issues and/or if there is a newer firmware level. If the firmware is very new you might actually try a previous one.

 

[Yossaris]

I had a continuous ping running on 2 machines. 1 in the usual subnet and 1 in the test subnet and both worked the whole time. Only the smb-connections seem to be affected.
I suspect that somehow the smb connections don't get closed properly and "overflow" for a lack of better term. I will keep an eye on the smbstatus of the NAS and try to get some kind of information out of it. (I'm relatively new to Samba, so I don't know if that makes sense)
There is an even newer version of the DSM for the NAS that I will also apply sometime soon, maybe that will fix it by itself.

 

[Ralston18]

Regarding:

"Synology NAS Cluster (DS1817 - DSM 7.2.1-69057 Update 6) shared folders suddenly become inaccessible after some time. That period can range from days to hours and I think it depends on the amount of traffic in the subnet in that time."

Is it correct that sooner or later those shared folders become available agaIn?

How are you being notified when a shared folder becomes inaccessible?

 

[Yossaris]

As far as I am aware, they do not become accessible again on their own. I start a switchover between the active and the passive server and that "fixes" it for a while. Like I said, I cannot feasibly wait a very long time before people get upset because they cannot work and I cannot replicate the error whenever I want, so I can't wait for the weekend or the end of the "normal" workday unfortunately.

I tried setting up a sensor to monitor the shared folder but it doesn't seem to work, so I need to wait for a user to notify me that the folder is not accessible.

 

[Ralston18]

Not sure about "sensor" per se. However, there are many things you can do to watch for precursors to the folder(s) becoming unaccessible. E.g. network pings to determine if some device drops out.

The folder disappearances are likely some "perfect storm" of events. Event A = ok, Event B = ok, Event C = ok.

All three Events together = Not ok.

= = = =

On the Synology NAS open the Control Panel and take a close look at all File Sharing configuration settings.

Overall quite complex and many details involved.

Note the small upward or downward pointing " > " that expand or contract other available settings.

Compare folders that do not disappear to folders that do disappear (with or without reappearing).

Compare end user rights and configuration settings: those users who report the folders disappearing to users who do not experience folder disappearences. Subnets, IP addresses being used, pathnames, etc.. access priviledges.

Look at all of the logs. Router and NAS.

Set notifications/emails. Honestly not sure about any specific rule or flag to be used.

Just set a few that seem applicable to your NAS/Network environment and the folder sharing requirements.

Objective simply to obtain some notification about any changes within the NAS and/or network that may preceed the disappearance of the shared folders. Which may or may not be due to NAS.....

I still suggest that you sketch out a full network diagram showing all network device along with their names, MACs, IP address (Static or DHCP), and subnet net mask. No duplicates.

The diagram should represent the physical network topology and account for every device, port, cable. And wireless as well.

The diagram may or may not take a lot of work. Especially with respect to the described subnetting. How many network devices overall and how many in each subnet? What is the IP address range for each subnet?

When devices drop from a network the first thought is that there is some physical cause. However there can be other harder to discover reasons. Duplicate IP addresses, access rights conflicts, some automated action that conflicts. More devices than allowed by the subnet....

Start keeping track of who reports the files have once again dissappeared. What, when, etc. Who, what else is online and connected to the shared folders. Dates times. One folder or all.

Eventually it will narrow down to maybe a specific subnet and then further down to some specific devices or users. Maybe someone has brought in a rogue device for their own purposes.... I would look for such.

You might find the root problem and solve it all. Or you might fix one thing to have the problem reappear because Fix A broke something else.

The described problem is simply one of those that takes time and effort to grind through. And may likely need some late nights and long weekends.... 🙁

Having the big picture to view is very helpful.

And if the diagram is going to be a difficult task. Look for and download a trial/free version of some network software that will do discovery. No harm in trying that.