Network drops every nite within 2 hour window
W2k3 Server runs recreation application. Server connected to Netgear switch which is connected to entire Netgear switched environment using fiber connections. Every nite between 6:30 and 8:30pm, the users using the application experiencing a lockup or server error. Event viewer shows hanging application on client pc but the server has nothing indicating network drop. I think it may be a switch issue and was planning to reset the power on each switch that is connected. The enduser makes 3 hops to server.
I also have Arcserve running nightly backups around that time. That is the only scheduled event that I am aware of running.
Leads me to believe since it is a timed failure that it is the tape backup causing congestion and possible failure,
or
its a switch that coincidentally fails every night at the same time,
or
we have been infected by something and it is transmitting something out of network on the scheduled time.
What is the recommendation or starting point to solve an issue like this?
I also have Arcserve running nightly backups around that time. That is the only scheduled event that I am aware of running.
Leads me to believe since it is a timed failure that it is the tape backup causing congestion and possible failure,
or
its a switch that coincidentally fails every night at the same time,
or
we have been infected by something and it is transmitting something out of network on the scheduled time.
What is the recommendation or starting point to solve an issue like this?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
I agree that I dont think the switches would react every nite at about the same time. The client PC is at the rec center switched from there via fiber module to switch at PD then travels again over fiber to the server location. I say 3 hops but they are not router hops just switching.
One of the backups does go across the fiber to the PD to backup a server at that location but the timing just doesnt seem right. This traffic happens hours after the client PC experiences the dropped connection.
One of the backups does go across the fiber to the PD to backup a server at that location but the timing just doesnt seem right. This traffic happens hours after the client PC experiences the dropped connection.
Make sure there isn't a time based ACL applied to this PC.
ASKER
Moving the backup forward 2 hours resolved the network bottleneck assume to be causing the lost connection issue. Short term solution.
If you use managable switch then you can enable SNMP and use Cacti or MRTG to check port utilization.
http://www.disorder.com/~bsod/Cacti-0.8.7i.exe
Moreover you can use Wireshark Packet analyzer to analyze packet at that time.But your switch should have to support Port Mirroing:
http://www.wireshark.org/download.html
You can use KS Host Monitor Software which could monitor all devices and if any fails it'd generate alarm/email or notification.There are so many monioring tools available online can perform the same.It'd be better for you to find out which device is causing problem.