Problems with 2003 NAS

[ehstech]

This problem has been plaguing us for some time. We are a high school which has a NAS which contains most of our data shares including staff, students, and our grading program. We do folder redirection for My Documents so the server gets used quite a bit. Last year, we had a 725N NAS which had been running fine. Then, all of a sudden, it started freezing every few weeks (you could ping it, but it would not reconnect shares; for example, if a user tried to open my computer to access a mapped network drive, explorer would freeze on their computer). At first you could generally access the server through remote desktop but after a few minutes that session would disconnect leaving you no choice but to do a hard restart by holding in the power button. We went through troubleshooting with Dell and they pretty much sent us all new guts of the server, but we still had the problem. So, we retired that NAS and bought a 745N.

After installation of that NAS, we are again seeing the same problem. We have tried different cables and a different connection to our network. In fact, right now I have load sharing between the 2 GB NICs enabled going to 2 different switches. Nothing ever appears in the event logs. The problem always occurs during the school day (when it is most in use), but otherwise there are no predictable patterns. Also, periodically, if you click on “sessions” under shared folders in computer management, that application will slowly trickle in user names (maybe 1 every 3-4 seconds) and will eventually freeze that instance of computer management.

Does anyone have any ideas of what else to look at? As you can imagine, we have some pretty unhappy students and staff when they keep loosing their work every few weeks because of this! Thanks in advance for your help.

 

[Law]

What about the switch you're using, brand and model? Are you using NetBIOS as your IP to name resolution?

 

[ehstech]

Thanks for your reply. We have a domain controller which runs DNS. However, most of the time, connections are made by IP address as in all of our group policies which apply folder redirection, map drives, etcÂ…, it is the IP references, not the name.

The NAS was previously connected to a Cisco 2950. Right now, it is load sharing between a Cisco 3550 (which acts like our router) and an Extreme Summit 200.

 

[Law]

Could you point me to a link for the specific Dell Powervault 745N NAS with the hardware spec that matches yours? I tried googling it but there’s a lot of different hardware spec. Can you estimate during the “peak” time how much students are accessing the NAS, because I think it might be too many for that one to handle. I just want to know the hardware spec so I can be sure. I can’t really say there’s anything that might be wrong with the 3550, I have work with it and it’s a really powerful switch but I am unsure about the Summit 200.

 

[ehstech]

Hard to give you a link to our exact specs since apparently Dell doesnÂ’t make this anymore, but here is the gist of the HW from DellÂ’s site with our service tag:

ASSEMBLY, CABLE, Serial ATA, COST EFFECTIVE RAID CONTROLLER, HW-RAID

PROCESSOR, 80546, 3.0, 1M, PENTIUM 4 PRESCOTT DT, 800, E0

CARD (CIRCUIT), CONTROLLER, Serial ATA, CERC-SATA-6CH

ASSEMBLY, TRAY, MOTHERBOARD, QUANTA, PE750

2X DUAL IN-LINE MEMORY MODULE, 512, 400M, 64X72, 8K, ERROR CORRECTION CODE, 184

4X HARD DRIVE, 160G, NATIVE COMMAND QUEUEING, 7.2K, 8M, LEAD FREE, MXT-SA

CARD (CIRCUIT), CONTROLLER, ADAPTER, 39160, FW3.10, U160M

For peak usage, I would estimate at most 200 users (between staff and students).

At one time, we had this plugged into just the 3550, with the same (freezing) results.

Maybe we do have too much going to it (this has been one of my fears), but I am puzzled because we have another high school in the district with just as much traffic going to their NAS (and their specs are actually lower than ours as they match the first NAS we got rid of), and they have no problems. I also am puzzled why I have problems viewing the sessions in the summer when there are only a few users on it...

Thanks for your help!!!

 

[Law]

I would say 1 gig of RAM is really low for a NAS server with high traffic volume going to it. It might not be the problem but that’s just my thought. For some reason I keep having this déjà vu like I have troubleshoot this problem before somewhere and the problem was in fact with the switch. I’m just not sure what exactly is wrong with the switch.

 

[ehstech]

Thanks for all your responses, Law.

Thanks for the point about memory-I have looked at it a few times throughout the year and the memory usage has never seemed that high, but it is still worth watching further.

If you ever think of the problem you might of encountered similar to this, please let me know!!!

 

[technjeep]

EHSTECH - we are experiencing similar no same problem with 745N - only we have 2GB RAM - about 45 users. Lockups are random although they seem more frequent in morning. Company moved 4 days ago to new location and had a suspect physical infrastrucutre including sub-standard wiring, daisy chained switches throughout the old building, etc. New location is all CAT6 certified with new Dell Powerconnect switches centrally located. Only one NIC active on NAS. Thought we would wait and see if old infrastructure was causing problem, obviosly not as we experienced our first lock-up today. We have been scratching our heads on this one as the event logs are empty, and the rest of the network is clean. Will be watching here and let you know if/what we come up with. Thanks.

 

[ehstech]

Thanks technjeep---does sound like the exact same problem! Now if we only had a solution...Let us know if you come up with anything, I will do the same.

 

[technjeep]

I ran performance monitoring for over a week and after review noticed that the paging file pegged at 100% utilization during the periods of lockup. I am going to adjust the paging file and will report back.