7.2-STABLE -- processes stalling

P

pterodyne

Sorry for the lame post, but I have no where else to go. Our IT staff and the entire company is almost completely gone. Our mail server is running 7.2-STABLE, and has been just that. Stable as possible. In the past week processes have begun stalling. For example. A 2 meg attachment will begin do download for a user to outlook via UW-IMAP. The process will never complete, or if it does, it takes 30 or 40 minutes.

This server at one time handled over 200 active users (with no performance issues). We are now less than 20. There are a couple of really small websites that live on this server too, and they stall when loading most of the time. (not all the images appear to download.)

Things I have done:
rebooted, rebooted, rebooted. Helps for about 15 minutes.
Investigated possible DOS attack with t-1 provider
Removed the gmirror (the second disk wasn't there anyway), pointed fstab back to /dev/asd1
Searched google and here. Don't know enough to even know what to ask.

Here is the output of vmstat -i dont know if that is useful or not. Again, sorry for the uninformed post, the guy that handled fbsd is no longer available, and all I generally have done is add and remove user accounts/email setups.

Code: 
vmstat -i
interrupt                          total       rate
irq1: atkbd0                          70          0
irq14: ata0                        68744          8
irq23: uhci0 uhci+                     1          0
cpu0: timer                     15383082       1911
irq256: em0                       129229         16
cpu1: timer                     15382909       1911
cpu2: timer                     15346859       1906
cpu3: timer                     15346858       1906
Total                           61657752       7660

Thanks.
 
Additional info:

I just saw this in dmesg.. Hadn't seen it before:

Code: 
ad0: 953868MB <Seagate ST31000340AS SD1A> at ata0-master SATA150
lapic1: Forcing LINT1 to edge trigger
SMP: AP CPU #1 Launched!
lapic3: Forcing LINT1 to edge trigger
SMP: AP CPU #3 Launched!
lapic2: Forcing LINT1 to edge trigger
SMP: AP CPU #2 Launched!
Trying to mount root from ufs:/dev/ad0s1a
em0: link state changed to UP
 
I now have the install running on a different machine. Seems to be behaving the exact same way.

The load average stays pretty low, .04 right now and dropping..
 
The hard drive went back in the original machine last night since it is quite a bit faster, even with the problem.

Our T-1 provider cant find any issues with connectivity.
 
Turns out, the problem was a dying switch. Never occurred to me since it is a fairly new, really expensive managed gigabit switch and there are other servers on there (I had tried different ports and cables already). This morning I tried to access a different server on this switch, and the same problem was happening. Services up and responding, but almost no data going through.

Well at any rate, sorry to waste anyone's time who might have read these posts.
 
Not at all. If this means people with network problems look at their switches before complaining about the FreeBSD network stack, I'm all for it ;)
 
You must log in or register to reply here.
Back
Top