PowerConnect 6224 6248 firmware issues

Question

Has anyone ran accross any firmware bugs in firmware 1.0.0.27.

Here are the bugs I have run into and how to aviod them:

1)PowerConnect 6248 switches reload sometimes when trying to run Statistics/RMON > Charts > Ports Statistics - Received Rate or other settings.

I put a case in with dell about this and here was their response:

"The RMON problem you encountered is an issue we've identified as a memory leak. There is currently no workaround, aside from simply not using RMON statistics. This is being addressed in a maintenance patch that should be released shortly. I don't have a timeframe on this, but this is the foremost concern for the new firmware's release."

Avoid this bug by not using RMON

Please help me and others to avoid future problems by posting other bugs here. Dell did breifly release firmware 1.0.2.4 and it was suppose to fix the bugs listed below. But as far as I know they pulled the firmware because it contained some new bugs which could cause the switch to crash.

I hope they come out with a new verison soon. I am very happy with these switches based on cost, but am starting to worry about their quality. I think they really need to make any known bugs public knowledge to protect their customers.

Bugs that probably exist in 1.0.0.27:

Title Description ID

Drawing RMON Charts

causes switch to crash

RMON leaks memory causing crash. Rmon task leaks around 48 bytes of memory

every second. There was no matching free() function for a malloc().

60555

Jumbo frames not passing
over LACP.

Jumbo frames not passing over LACP. Jumbo frames of size more than 1518 bytes

are not forwarded onto a lag

59739

Cannot disable GVRP via
SNMP.

Cannot disable GVRP via SNMP. When trying to disable GVRP via SNMP in the

dot1dBridge->dot1qBase MIB Object, the SNMP returns a failure. If disabled via the

CLI, the SNMP can enable it OK, but cannot disable. The CLI can enable & disable

OK.

56695

Arp requestprimary/
Secondary

When sending ARP request, select srcIP in same subnet as targetIP. The issue as

described that the IP stack was always using it's main IP address rather than 1st

checking if there is a secondary address match on the subnet where the ARP request

was being sent on

59237

Random failures when the
stack size is large Example
stack of 9 units

Random failures when the stack size is large Example stack of 9 units. There is

sometimes a loss of communications from a master to its members, although the

master detects the presence of the other members at some level in the protocol stack,

control/status information is not being freely transferred.

59207

cannot delete startup-config
This issue is that it is possible that Flash File system when it detects a corruption

which could result in the powering down of a device during a write operation to flash,

the FFS driver will mark the FFS as read-only. This will result in end user

experiences that such operations as "cannot delete startup-config" will result. File

logging has been set to default off.

58639

Device goes down when
"dot1x system-auth-control"
command is issued

dot1x timeout re-authperiod has a range of 300-4294967295 but the corresponding

SNMP variable cant configure more than 65535

61149

WEB-UI does not report port
status

WEB-UI does not report port status properly 60674

CLI-show vlan-doesn't report
all members of VLAN

CLI-show vlan-doesn't report all members of VLAN 58629

Counters not getting updated
correctly

Counters not getting updated correctly in case of Jumbo Frames 59739

With Persistent logging disabled, output of show running-config is wrong 61143

Not able to backup running
config to remote location via
SNMP

Not able to backup running config to remote location via SNMP 60491

Display of Multiple VLANs on
"Show Running Config"

Display of Multiple VLANs on "Show Running Config" 60381

switchport forbidden vlan
remove command does not
work

switch port forbidden vlan remove command does not work 60173

input Port Description using
GUI, then port goes down

If you input Port Description using GUI, then port goes down 61559

62xx device in port 25 causes
STP reconvergence

60480

Flash Diagnostics usability
In the Run Flash Diagnostics menu, if you enter ? in the iterations it will run 99 times 67720 global status does no t
change on 62xx

stopped fan to produce a global status change, status does not change 62911

jpugsley · Answer

I'd like to add an item, and it's difficult to tell if it's a bug or just some arbitrary limit on the system.   ACL - mentions 1024 Lists (Limit) There is a limit of 10 Rules per ACL.   I'd ask if anyone else is finding this issue of 10 Rules per ACL to be too restrictive. I myself need to create an edge list that contains minimally 74 Rules. I was told by Dell Technical Support that when this new firmware is released it will allow us to apply an unlimited number of ACL's to an interface. Which in my case would require the create of 8 separate ACL's. Apparently it's been coded so that the last ACL that is processed is the only one that will contain an Deny ANY ANY statement.   So, if you agree that the 10 Rule limit on ACL's pretty much makes it impractical to use please let Dell know.

ev89pmp · Answer

Hello all,       I am having many problems with the 1.0.2.4 patch as well. In my situation, if I attempt to goto the interface configuration screen for the Routing function, and the switch died a slow death only to come back to life in about 90 seconds... If you are reading this and you have NOT installed version 1.0.2.4.... DO NOT DO IT!     Elvis

Adam N · Answer

Hi,   We are about to release firmware 1.0.2.6 in the coming days which will resolve the bugs mentioned on this page as well as the fatal error routing bug in 1.0.2.4.    I will post an anouncement on this forum when the new firmware has been released.   Thanks

ev89pmp · Answer

Thank you!   Elvis

stormytoad · Answer

1.0.2.6 firmware has been released.  I've loaded it on several 62xx switches and it seems to have fixed the major issues I was having.  I'll post here if any new problems come up.

padman37 · Answer

I am unable to tftp the new firmware onto one of our new 6248's. 'File Transfer Failed.' Is it because it is in the default factory config of a master 1 stack? I remember somewhere that we had to take out our 3448's out of a stack before we can upload a new firmware. If so, how do I remove the stack and set it to single mode? Anything else it could be? I have tried numberous tftp programs with the same failure results. What am I missing here?

pauljvsg · Answer

I was unable to get 1.0.26 to load via tftp or xmodem. I had two brand new switches that were stand alones. Would not load on either. TFTP tranfers just failed or would fail crc after tftp transfer. Xmodem would just quit about 1/3 of the way through. I had to verify flash (or whatever it says on the boot menu), this deleted evreyhting and I was able to load via xmodem after reentering the mac, service tag and sn. I changed the baud rate to 115xxx to speed up the xmodem trans, it should load in about an hour.

Flynder · Answer

I had the same problems upgrading the firmware of 2 new 6248's from 1.0.0.27 to 1.0.26

I had _no_ luck with CLI-tftp, Webinterface-tftp or Xmodem transfer

**** Warning this fix erases your existing firmware and the switch WILL NOT BOOT until you succesfully transfer a new firmware ****

1. Attach a serial cable and reboot into Boot Mode [2]
2. Select "Run Flash Diagnostics" :emotion-14:
3. Choose 1 iteration and drink a cup of coofee while it erases the flash card
4. Reset the switch [9]
5. The switch now has no firmware - you might have to enter the Serial number, Service Tag and MAC address - one out of the 2 switches required this information.
6. Change the baud rate to the max 115200 [2]
7. Reconnect at the new baud rate
8. Load the new operational code ie. firmware - takes about an hour [4]
9. Update boot code [7]
10 Reset the switch and you have now upgraded the firmware [9]

Networking General

PowerConnect 6224 6248 firmware issues

Was this post helpful?