Networking issue

[autumn]

G day all

I have an issue with a program which is build on MS access. It's needs constant connection to the server via a mapped drive but for some reason on the equipment with the low power chipsets the connection looks like it drops and the program/database cracks it saying it can't find this and that until you close it and reopen it. This is happening on a mix of Win 7, 8 and 8.1. The equipment are 1 Toshie laptop, 2 Dell convertibles, 1 NuC. Standard desktops are not effected.

I have already done the windows power management of USB and Network, wired and wireless, BIOS power management, ping -t server IP address, turned firewall off for Domain.

I was thinking that it was a switch as it's from one room but there is only one switch for the site.

The software company have washed their hands of it.

I'm at a loss at the moment.

 

[YeOldeStonecat]

Is this happening via wireless?
Or are you also connecting via ethernet cable and it still happens?

Does the Toshi have an Atheros NIC? If so...I walk away from them, and I tell the customer to get a lappy with Intel NICs. Wasted too many hours of my life years ago trying to troubleshoot and support Toshiba lappies with that damn Atheros wireless chipset. They don't handle loads well.

 

[autumn]

not to sure on the type on card.

It is happening with both wired and wireless. More so with wired but that may the user use. it seems to be happening when they aren't using the program for about 30 minutes. The Dells are using USB dongles or Docks via USB. The Toshie is wireless but does it while wired. and the NuC is wired

 

[YeOldeStonecat]

Hmmm...strange. Guess the common denominator here is "alternate" NIC connections. Meaning not a nice onboard Intel NIC....but like you mentioned..via USB dongles, or via docking station (which often are still just a funky type of USB port under those bomb bay doors).

That's an odd one. All I can think of is where you've already gone...deep into the BIOS, and deep into the Winders power management settings..even into that advanced section.

Maybe power management isn't the issue? Does this happen after a period of inactivity? Or does this happen during use?

 

[autumn]

yes it does seem like an inactivity thing but it can be less than 1/2 hour.

I did think it was something to do with the USB dongles or Win 8/8.1 as the Dell's were the only one's with the problem more so with 8.1. Then I put the Toshie in with win 7 and that started to have the same issue, and then they got the NuC with w8.1 (which I wouldn't put in until I had a solution) one of the reasons I put the Toshie in with win7.

I'm on site tomorrow and check the network cards.

 

[autumn]

OK, it doesn't seem to be the nic as there's nothing common between the three types.

Dell's Centrino advance n6235
Toshie Atheros ar8171/8175
NuC intel i218v

going to setup a SNMP trap over the weekend.

Any other suggestions??

Regards

Tim

 

[pctechsupport]

Hi there

A few things that spring to mind that I would check but no guarantee any of these will help. Maybe you have checked these already?
1. Is there AV on the machine? If so disable and test.
2. If AV is on the machine - check if there is an NDIS filter enabled in the properties of the network card (where under client for microsoft network). If there is disable it?
3. Ensure the MS Access Trust settings are correct and MS Access trusts the server/network.
4. Can you access the mapped drive when the issue happens?
Check if there may be a setting related to the length of time the mapped drive stays alive?
5. Do they all disconnect at the same time? If so could it be related to inactivity between the MS Access and the server or the shared drive.
Test turning them all on at the same time and connecting MS Access to server. See do they all disconnect then at similar time.
6. Check that its not the case there is a limit on the number of connections to server/MS Access DB/Shared Drive
7. Can you ping server by name and ip from the machines when the issue happens?
8. Ensure wired and wireless cards aren't in use at the same time on each machine. Ensure whichever one not in use is disabled.
Ensure network drivers are up-to-date on all cards.
9. Ensure the lease time for DHCP is a few days and not hours incase its impacting it?
10. Ensure all machines are clean of viruses incase there is a rogue virus on 1 machine causing network broadcasts clogging the network?
11. Ensure there are no other switches/access points/hubs in place that you might not know about.

 

[YeOldeStonecat]

Dell's Centrino advance n6235
Toshie Atheros ar8171/8175
NuC intel i218v

If it were just the Toshi I'd blame that and throw it out...as I hate hate HATE Atheros wireless chipsets. Man..the amount of time I've spent wrestling with those...will flat out tell clients I can't guarantee them anymore.

But you have 2x others in the mix...so there's still some issue.

So how about swapping out the cabling?
Perhaps tossing in a loaner/test switch....maybe the switch is faulty.

 

[autumn]

Hi there

A few things that spring to mind that I would check but no guarantee any of these will help. Maybe you have checked these already?
1. Is there AV on the machine? If so disable and test.
2. If AV is on the machine - check if there is an NDIS filter enabled in the properties of the network card (where under client for microsoft network). If there is disable it?
3. Ensure the MS Access Trust settings are correct and MS Access trusts the server/network.
4. Can you access the mapped drive when the issue happens?
Check if there may be a setting related to the length of time the mapped drive stays alive?
5. Do they all disconnect at the same time? If so could it be related to inactivity between the MS Access and the server or the shared drive.
Test turning them all on at the same time and connecting MS Access to server. See do they all disconnect then at similar time.
6. Check that its not the case there is a limit on the number of connections to server/MS Access DB/Shared Drive
7. Can you ping server by name and ip from the machines when the issue happens?
8. Ensure wired and wireless cards aren't in use at the same time on each machine. Ensure whichever one not in use is disabled.
Ensure network drivers are up-to-date on all cards.
9. Ensure the lease time for DHCP is a few days and not hours incase its impacting it?
10. Ensure all machines are clean of viruses incase there is a rogue virus on 1 machine causing network broadcasts clogging the network?
11. Ensure there are no other switches/access points/hubs in place that you might not know about.

OK as follows

1: yes and done
2: no NDIS
3: this is all controlled (locked down) by the software company, if you didn't know you nearly wouldn't know that it's an Access db.
4: this is interesting, the simple answer is no but one of the clients does show as disconnected some times.
5: no different times and different lengths of time and the problem is only on the mobile devices, as such, the desktops are fine.
6: connection numbers shouldn't matter but will check with software company
7: Yes ping is replying fine and I have tried running ping constantly but the problem still exists
8: yes only one is being used at a time, and up to date
9: DHCP time is 3 days
10: no viruses that is know
11: yes was thinking a switch problem, I have tried to rule it out by swapping the ports with the desktops with n change, they have digital phones but have isolated them as well.

Thank you.

 

[autumn]

If it were just the Toshi I'd blame that and throw it out...as I hate hate HATE Atheros wireless chipsets. Man..the amount of time I've spent wrestling with those...will flat out tell clients I can't guarantee them anymore.

But you have 2x others in the mix...so there's still some issue.

So how about swapping out the cabling?
Perhaps tossing in a loaner/test switch....maybe the switch is faulty.

Yes agree, I normally supply the hardware to this customer but all of this they have brought elsewhere so full chargeable as well. False economy for them as I don't charge install or time on warranty if brought from me but, they save $100 on the hardware but I charge $200 to install. Anyway.

yes if they were all simular nic's then I would have said that but no.

I've tried cables and changing ports in the switch to no avail. yes have been thinking about swapping out the switch.

Thank you

 

[autumn]

ok found out more today.

It's not a timing thing.

it doesn't happen to all at the same time.

It happens while they are using it.

network drives are still active

will be changing switch next week

 

[pctechsupport]

Good idea to change the switch as a test.

Also if you have time - download and install Microsoft Network Monitor (or Wireshark) on laptop that has the issue and a desktop that doesn't. Run the monitor at the same time on both and then once the issue occurs review the packet trace to see if you can identify what the difference is for traffic specifically going to and from the server?

I would check the application and system event logs on the server and the affected laptop(s) and see if you can see any errors around the time the issue starts that might point you in the direction of the issue?

Also probably nothing here but might be worth checking task scheduler on the laptop(s) to ensure at the time of the issue a task that might have kicked off isn't affecting it.