• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Networking Pause frames

Y

yvsbenchmark

Junior Member
I was browsing through internet for a situation that is happening in my setup that the SAN becomes slow everyday, this happens even though my MetaData Server is having normal CPU and network performance.

Equipments being used in the setup are as follows 1. SuperMicro Switch Layer 3. 2. Tyrone SAN: Connected to Switch using Fiber 10G 3. MetaData Server(RAM 32GB with two Intel Xeon Processors): Connected to Switch using Fiber 10G 3. Editing Systems are FCP on MAC: Connected to switch using 1G ethernet

Scenario:

Usually when we test the SAN it works file but everyday at some point of time the SAN starts slowing down. and editors are not even able to play clips on SAN and we start getting the beach ball if we open the SAN on MACs and slows down the complete network.

OBSERVATIONS:

MDC network usage is not more than 4% - 5%
MDC CPU usage also goes MAX upto 50% - 60 %
When checked on switch we find the following observations


PORT___MTU___SPEED__RX Octets__RX Ucast__RX.Bcast__RX Mcast__Rx Pause__RX Discard Ex0/4___1500___10Gbps_1963198694_289864464_1832033_31552____11434_____45636
Sorry for ugly representation of statistics but this is because I cant post an image.

the above statistics are after I restarted the switch yesterday only.just to check if the pause frames are still noticed.

Can anyone please guide me if this can be the reason of my SAN slowness? and if yes how can i take measures to prevent it.

Thanks in Advance. .

Regards, Yash
 
This reeks of some job being kicked off at a specific time each day, does the SAN have any built in maintenance tasks? AV scans? Backups?
 
Thanks for the reply. I checked the complete setup but there was no task running in the SAN or any machine connected to it. I have even removed antivirus from all the machines connected to the storage.

Regards,
Yash
 
yes usually the work load is at the max at that point of time but it is just about 3-4 macs editing over the storage which overall is not high.. secondly sometimes when there is like only one mac editing over the storage, it gets slow.
 
Im grabbing at straws since i am not familiar with your particular model of SAN, but does it have any sort of performance metrics that you can consult? Im not entirely convinced this is network related although i cannot rule it out at this time. A few gb nics should not be able to saturate a 10G link
 
Thanks a lot for the reply. The SAN is Tyrone and it is an Iscsi based model. the thing that makes me think about network is that at the time of slowness the MDCs are not even using 15% of their network and 40% of their CPUs. while the pause frames and discard frames on the switch keep on growing in the web interface.

Regards,
Yash
 
Do you mean a flowcontrol pause frame? The SAN can send those when disks are overloaded as can any other flowcontrol enabled interface.
 
PORT___MTU___SPEED__RX Octets__RX Ucast__RX.Bcast__RX Mcast__Rx Pause__RX Discard Ex0/4___1500___10Gbps_1963198694_289864464_1832033_315 52____11434_____45636

This is the detail of switch which i find in the swith GUI. the SAN is 40% full.
 
you should not have but 1 or 2 discards per day if flow control is working properly under heavy ISCSI usage!
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top