Having the same issue on PS4000 running firmware version 5.0.5.
We have updated the switches to the latest firmware and are working with Dell on the issue.
Hopefully we will see a resolution soon.
Did the update v5.0.5 work for anyone?
Opened case with Dell, fyi
Dell following up - We are hoping to have this issue resolved with an upcoming firmware release. Dell wanted to check that the speed & duplex settings on the switch ports connecting to the array have not been changed from anything other than automatic.
me- Correct. Would that be firmware update 5.0.6?
Dell - At this time we don’t have an exact release date or revision for the firmware that will resolve this issue. However, once it is out, we will let you know.
Firmware 5.07 just released. If anyone updates his/her EQL please let us know if it was fixed.
I do not believe this issue has been fixed in 5.0.7 as it would have been placed in the release notes, however, the issue has already been confirmed by support as not a "problem", as it was posted here. I will be upgrading my arrays to 5.0.7 this coming weekend and I will post back if this issue persists, but right now I haven't recieved any more errors on the one PS4000 that had the errors, all other arrays never exhibited the problem.
I recommend opening a ticket with EQL support if you are still seeing a high number of packet errors, just to confirm nothing else is going on in your arrays.
I got conformation from EQL that they have found the cause of the problem and that it should be solved in the next firmware release (5.08?)
According 2 Dell technical support (haven’t tried it yet!!);
Packet errors on PS4000E connected to PowerConnect 6224 switches; a firmware patch V5.0.7-H1 that can be applied to your system as a resolution for this issue. To download the patch, login to the EqualLogic support site first, then paste the following URL into your browser: https://www.equallogic.com/support/download.aspx?id=10559
You upload & apply the firmware patch in the normal way. Applying this patch will require a controller restart, so iSCSI connections may be affected. We recommend that this patch be applied when downtime on the hosts can be arranged as a precaution.
IT's even more intrigued on my site.
I have (on one site) 8 ESXi whcih run 3.5i without any EQL issues 9except dumb multipath which is not compatible with anything). No errors.
We upgraded 2 to ESX 4.0.1 (not 4.1 because we need to do it in stages) last week - no errors.
Then I upgraded 3 more to 4.0.1 yesterday and all 3 reports iSCSI errors causing reconnection every hour or so.
So, it is definitely something in both EQL (because not erors with other iSCSI arrays) and ESX4i (because no errors on previous iSCSI stack in 3.5i - 3.5i runs kind of cisco iscsi and 4 runs something which looks as open-iscsi).
I'll post update if manage to resolve this problem.
Equipment is - dell R710 and dell 1950 (one of last 3 systems and one of first systems are dell 29xx or 19xx, the rest is R710), original NIC-s, cisco switches; not sure about jumbo (EQL was configured long ago and worked pretty well aside of multipath problems which was expected).
(I think that some iSCSI or TCP optimization in ESX4 broke iSCSI with EQL under the load bursts).
What's interesting too - no visible impacts on performance.
Have applied firmware 5.0.7 and patch V5.0.7-H1 (V5.0.7 (R178545) H1) and no more errors.