SG300 series. Flashing lights, Poor Throughput, Disconnect/Connect

[csparrow17]

Hi all

We had a number of Linksys SRW2048 / SRW2024 switches that had the "bloated capacitor issue" or "bad capacitor issue".

These units would basically disconnect/connect their ports.

Applications would sometimes crash on workstations.

These were all located at different client sites.

These were replaced under warranty with the SG300-52 units.

When we implemented the SG300-52 units at several sites we noticed the the logs showed disconnect/connects.

We were not seeing apps crash.

Support advised they were not aware of any issues.

We did a firmware update that was available that made no difference. (1.1.0.73)

6 months have now passed.

We noticed at our site very poor throughput one night doing a very heavy file copy data transfer between two computers.

It actually ground all other computers on the network to a halt.

We inspected the switch.

It had "all lights flashing".

We thought this strange and shutdown the whole system. Problem still existed.

Power cycled the switch. Problem still existed.

Noticed their was a new firmware update. (1.1.2.0)

We implemented this.

The lights on the switch stopped "all lights flashing".

Throughput was back to normal. (ie didnt grind all other computers to a halt)

We then had a second site complain of similar issues.

Heavy throughput between two computers bought all other computers to a grinding halt.

Lights were "all flashing" on switch.

Workstation would disconnect/connect in a loop during heavy file copy.

We implemented the firmware update thinking this would solve the issue.

It now still has the issue of disconnect/connect however and apps sometime crash.

We have now replaced the switch with a HP 2510-48G.

This is working properly.

I have now since been to other sites that also have these SG300 switches and notice the "all flashing lights".

These model switches are a serious concern...

[rocater]

Hello Chris,

The issue you describe is very similar to a problem that was fixed in the latest firmware. The issue was that after a period of uptime, 6 weeks or geater, the switch would stop learning mac addresses. This would cause all ports to light as the switch would be acting more as a hub. If you still notice the issue after upgrading, please let us know.

[csparrow17]

Hi Robert

As mentioned above I am already running the latest firmware 1.1.2.0 dated 29 November 2011.

My issue at the moment is the disconnect/connect loop.

[jasbryan]

Chris,

I haven't noticed the behavior you have described; I have 3 SG300-28 ports in my lab with 4 NSS3xx and I’m replicating large amounts of data between two NSS3xx. I have about 40-50 hosts connected. Can you give me a better idea on the amount of data on your backups so I can send through for testing and see if i can accomplish what your switches behavior. I’m currently running a 75 GB file backup to another NSS3xx right now for testing.

All switches on 1.1.2.0

Jasbryan

[jasbryan]

Chris,

I ran 75 GB backup through my network  no symptoms. Backup started 15:21 to 16:49 1hr28min later. Since you're having trouble please give Cisco Small Business Support center a call and open a support case.  We can take a look at your configuration and assist. (1-866-606-1866)

Jasbryan

[csparrow17]

Hi jasbryan

The issue is now more the fact that the unit intermittantly disconnects then reconnects its ports in a loop.

This particular client runs a Microsoft SQL2008R2 database application.

This particular client application seems to crash whereas other apps like Outlook can recover from the fault.

In fact with Outlook you wouldnt even know it has happened.

You may not even have known you had the fault when you did your file copy unless you check the logs on the switch and your PC or you happen to see it on your PC screen.

When I had it fail I could see Windows 7 file copy recover from copying the 120GB of iTunes data.

I could also see the network icon in the system tray disconnect then reconnect in a loop.

(This was prior to the firmware 1.1.2.0 update).

I am now testing the unit in a partially non production environment.

It is currently working, although I notice it may have the "all lights flashing" issue again.

I also notice our production SG300-52 is also doing the same thing.

It has firmware 1.1.2.0

Will test a little then perhaps call support.

Realistically not sure what they are going to be able to do.

The information I have posted here needs to go up the chain...

FYI - Am in Australia.

[jasbryan]

Hi

No worries the Cisco Small Business Support Center is open 24/7 so please call at your next available time. When calling we need to have topology of your network, how you are planning on implementing these switches. We will take a look at your setting and verified or narrow down what is causing the switch to behavior as you are describing. Replication and testing has to be done first (trouble shooting) after we determine it's not a configuration or environmental issue we then can move your case up the chain.

Support numbers

Jasbryan

[nickjgrac5678]

good  day all I am having a similar issue in our network all the switches are all mainly SG300 and are all flashing continuously. Is there a possibility this could be a broadcast storm and if so what is the best way of finding the storm. I have tried wire shark but didn't really see anything too concerning and have also cleared all the counters on the switches. I would welcome some trouble shoot ideas one CCNA to another Help!!! at the moment nothing seems to be badly affected but it has been going on for some time and need some pointers in the right direction. Thanks in advance