Dell 2400 Server Problems

[rjs]

Has anyone had any bad experiences with Dell servers, 2400 PowerEdges in particular. We have been using PEs for awhile and were pleased. But we've had nothing but hardware failures on the 2400. I'm wondering if it is our bad luck or whether there is a more generic problem with that line of servers?<br>
<br>
Positive and negative feedback would be appreciated.

 

[Guest_imported]

We have 2 Dell 2400 Power Edge servers, and have had very little problems with them, the occasional software glitches, and I remember some problem with the RAID controller when we first set it up. It has been my experience, if we do have any kinds of problems with Dell hardware on any of the comuters (Workstations OR server) the tech support is always willing to bend over backward to solve our problem, or replace the faulty component...<br>
<br>
What type of hardware failures are you experiencing? <p>John Vogel<br><a href=mailto:johnvogel@homepage.com>johnvogel@homepage.com</a><br><a href=

http://i.winfire.com/s/isapiEng.dll/wf.exe?cmd=rl&452,20004878&wf.exe>Get

FreeDSL Service</a><br>[To a get FREE DSL Modem WITH FREE DSL Unlimited Internet Connection click the above link ]

 

[rjs]

The first problem was with the tape drive. They replaced that and the replacement failed. So they shipped out a completely new system. That new system failed within one week - we think we narrowed that down to the cache on the PERC/2 controller, which also resulted in not be able to recognize the existing RAID container. So all our software installed (NetWare, ARCserve, NDS, GroupWise, etc) was lost. We are on our second replacement 2400 - third server with our fingers and toes crossed.<br>
<br>
Dell tech support has been great, but their customer service stinks.

 

[PeteAppleton]

We have 2 2200's, 1 2300, and 1 2400.&nbsp;&nbsp;The only problem we've experienced was a failed drive in one of the 2200's.&nbsp;&nbsp;Both the 2300 (~1 yr old) and the 2400 (~2 months old) have been fine, although the 2400 isn't really old enough to expect any trouble.<br><br>I have had fun installing NT because of the necessity to hit F5 & install the PERC drivers /before/ NT prompts, but this was clearly covered in the documentation... when I read it...

 

[cyjae]

By now you have had a few replies about dell servers we have 4 two 6200's and 2 2450's I have had no problems with any of these servers. As such I have determined to ONLY purchase DELL servers, even though I won't purchase their workstations. But that is another story.

 

[mawheeler]

We had a very similar problem with our 2400 which came with a factory installed DDS4 drive. The tape drive would simply 'eat' tapes.

Eventually (2 weeks?) DELL supplied a new SCSI card specifically for the DDS drive which solved all the problems.

If this sounds familiar, I can get the details together if that would help.

Thanks,

Mark W

 

[mbarlam]

Admin for 30 Server Farm, I have at least 7 different Dell Models running here and they equally have certain problems. Mostly it's HD going bad. If RAID is enabled right, you're only scheduling a reboot. You're right about the F5 on NT install but if you use the Dell OpenManager Server CD, it does everything for you. I've installed NT in 20 minutes with that CD.

 

[kquinn]

We have just bought Dell PE2400 to replace older Compaq servers. From the beginning we have had a significant amount of hard drive faliures for servers shipped to remote sites. It seems these HDs are very sensitive to shock and we had to pack them better. Our other problem has been the Perc Raid contrller set at Raid 5 sometimes will not rebuild a filed drive.

 

[mbarlam]

some of the most recent Dell Workstations and Servers that I've been getting have been sluggish and problematic from the gate. My new practice is to reformat right out of the box. Whatever Dell is doing to the machines is effecting performance and reliability.

 

[vijmat]

We have around 30 Dell servers of different models. As a practice as soon as a box arrives, I use Dell OpenManager Server CD and rebuild the server fresh. We are in 4 locations and never had any server or harddisk crash. All our desktops are also Dell and we follow the same practice (putting our own builds) and we are satisfied with their tech support.

 

[Guest_imported]

DeLL PowerEdge 2400 RoX!

 

[aarau]

We had some trouble with Power Edge 4400 Servers. 2 Mainborads and 2 DLT's was at beginning out of order.
-------------------------------------------------------
Does somebody do uses the Raid Agents for these Servers
on NW51 ?
We have some problems if i start the Dell agents.
Dell has no idea!!!!!!!!!!! No help!!!!!!!!!!!!

Has somebody the Dellstart.ncf for a 4400 wit a pertec3
Raidcontroller?


Thanks
werner

 

[Grovenet]

I have had an interesting problem with a Dell Power Edge 2400 and backup drives. The solution was found to be the 3com card and Adaptec SCSI device as talked about in Microsoft Technet article Q153128.I could not find a satifactory result to the error except adding a different SCSI device to run the Dat drives and all errors went away as explained in the above article!

 

[SpecificWare]

We have three 2300, three 2400 and three 2500. They all have been working well. The only problem we ever experienced was when the write cache was enabled on the PERC controllers. Once this was disabled the servers are very stable.

 

[murraykr]

I have multiple Dell PowerEdge server models (14XXs 23XXs 24XXs 44XXs).
To put it kindly, I will never purchase a Dell server again!

Disks, tape drives, raid controllers, batteries, systemboards, backplanes, floppy drives have all caused problems (within first year of purchase).

Hard disks have been the worst culprit. In fact as I am writing this post I am on the 13th hour of a server rebuild after losing one disk in a raid 5 set!!!

Dell tech support has not been a lot of help for most of my problems, asking me to perform procedures on production boxes that are not sensible. Reboot here, reboot there, reboot everywhere!

Be prepared to be flashing/updating/upgrading/installing numerous patch/driver/firmware upgrades. it is extremely difficult to keep up!

I would be interested in hearing if anyone else has had multiple hard disk failures. It seems that a lot of the IBM hard disks have been failing. We purchased 6 - 4400 PowerEdge servers with 33 - 36 GB drives and have replaced at least a dozen of them since we purchased the servers a year ago!

 

[SteveTheGeek]

MrStress,
Sounds like the name fits. <grin>
I've dealt with roughly the same number of HDs over the same time period and haven't had one die yet. I wonder if there might not be an overriding cause for the failures, such as a high static-electricity environment? I once worked in a small IT training center (three 12-desk classrooms) that went through quite a bit of PC hardware until they provided true-ground copper wiring throughout all the classroom desks and connected the case of each workstation to that ground. Hardware replacement costs fell by something like 80%.
-Steve

 

[NickFerrar]

DLT drives have always been the weakness we've found in Dell servers. We have around 40 Dell servers and between them we've probably had 20 DLT drive replacements over the last couple of years (one system had 4 DLT drives replaced...).

Hard disks have always been very reliable for us, between those same 40 servers we've had only 3 or 4 disk replacements in the last 2 years.

Dell have recognised the DLT problem at last and now have a dedicated team to deal with it, unfortunately this just tends to delay getting support as they insist you run a lot more diagnostic tests (although sometimes it's useful).

The big weakness of Dell support is when soemthing isn't obvious. If it's a disk failure or something they usually have a replacement disk on the way to you within a few minutes (with an enigneer not far behind). However if it's something less obvious like backups failing with SCSI errors it can take ages to resolve and dell don't take ownership of the issue properly.

A few months ago the backups started failing on a server that had worked perfectly for months before (and nothing had changed on it). Initially Dell sent 2 replacement DLT drives, then a backplane and finally a motherboard. All that took over a month though - this was a critical production file server and I had to build a DR server to replace it as the swap-out and diagnostics downtime was far too great. Basically Dell support don't care how important the server is, they'll suggest rebooting as the first option 9 times out of 10. Also there should be an engineer that follows the problem through and is referred to if the call is re-opened. On our nightmare server I re-opened the call several times as the issue reoccurred and each time I got a new engineer who only had stuff like 'Swapped DLT drive' from previous calls to go on so insisted I run the same diagnostics I had run several times previously.

The firmware issue also irks me. We follow the general principle of 'if it ain't broke don't fix it' - so I don't go flashing BIOS's and ESM diagnostics just because Dell releases a new one. So if a system has been running fine for a year without change I don't expect to have to flash everything to the latest revision before Dell will even listen the issue I have - yet that's pretty much their standard policy (and we are a preferred account customer - God knows what help small businesses get).

One thing I don't follow about MrStress's post though is - if only 1 drive in a RAID5 set fails, how is taking you 13 hours to rebuild the server? It should just be stick the replacement drive in and do a consistency rebuild to populate the new disk - all while the server is on-line and usuable (albeit at degraded performance until the automatic rebuild is compelte)...

 

[1sexypapi]

we have many Dell servers 2300 24xx and 64xx there only issues that we have had with 24xx servers was one drive failure everything else good so far

 

[DanZeMan]

I work with MrStress, and although one would expect simply replacing a single failed disk in a RAID5 set would work - it didn't.
This happened on two different occassions, granted on one of them, a second disk failed during the rebuild (D'oh). But with only one dead disk, after replacing it, calls were coming into the HelpDesk with corrupted files messages. The entire RAID5 set had been corrupted. Now our SOP is to down the server to replace a &quot;hot swappable&quot; drive when it fails.
I wish that we could convince the powers that be to let us revert back to UNIX, at least that way, you get good, qualified tech support who understands what &quot;production server&quot; means, and on-site engineers instead of what we affectionately refer to as &quot;parts monkeys&quot;, who come in to replace a singe part, leaving you to go through the entire rig-a-marole again for the next parts swap.
As for the (very) high incidence of HDD failures, these are temperature, power controlled data centres, with full alarming on abnormal conditions, and logs of power issues, and we've had no indication of anything that could be causing these drives to die. The only curious things it that all the failing disks are IBM's, and we're getting the replacements as everything but IBM!