HighPoint RocketRaid 3522 Problems

[David Roth weiss]

Randall,

Its possible that the problem you’re having is with the G-raid mini, not the raid. I’m wondering if there’s a bottleneck and the raid is fouling-up because the G-raid simply can’t keep up.

Try xfering data to an internal drive as a test. If it works, you may find it necessary to copy to the internal drive and then to the G-raid mini.

Let me know if this theory has merit…

David

David Roth Weiss
Director/Editor
David Weiss Productions, Inc.
Los Angeles

POST-PRODUCTION WITHOUT THE USUAL INSANITY ™

A forum host of Creative COW’s Apple Final Cut Pro, Business & Marketing, and Indie Film & Documentary forums.

[Randall Brown]

Yea i tried that, it didnt help. I did the internal drive, the start-up drive, the g-drive raid, mini, and all the same. Whats really wierd, is that i can burn a DVD of the same files that wont copy when dragged. This has me stumped.

[Sean Oneil]

As long as write-back cache is enabled, and you have the latest drivers and firmware, then you’re doing everything right. It could just be a drive incompatibility. The disk drive makers rushed to get 1TB models on the shelves. I’ve heard a lot of problems with the Samsung ones. Haven’t heard anything about the Seagates, but anything’s possible.

Whenever you build your own RAID, it’s really important to make sure the controller card has been tested with the model hard drive you plan to use.

Do not troubleshoot this to death in hopes that it will eventually work. If it’s not to late, be prepared to return those disks and try a model that Highpoint recommends.

EDIT: I just checked and see that there’s a firmware update for your card released in January. See if you have the latest one. If not you gotta update it. Unfortunately you need a PC to do that.

Sean

[Randall Brown]

Today I am going to talk to Highpoint and see what they say. Luckily I sourced the drives locally for this very reason. I am going to see about getting a PC in here for the card update. Will post back, and if anyone else has this issue/ seen this let me know!

[Randall Brown]

So I think we have found the issue. It looks like the problem was the block size, it was set for 256k and for some reason was not working. It would not work at 128 either, it was only at 64k where it worked. Now I am sure I am giving up performance which is an annoyance, but the system has past the first few hang-ups that had crippled it and tests out in the high 300MB/s read and write (5 drive array uninitialized yet for performance)

A couple of things from this experience.

1. Seagate customer service was alright, had a Level 2 on the phone in 5 minutes, we went through all the drive potential issues, and came up clean. Their official final perspective though was that these drives dont have the error correction in them to operate in a group larger than 2 or 3 and they recommended i move to enterprise drives. Though they washed their hands of it, it didnt seem to be their issues since the drives worked out.

2. Highpoint tech support SUCKS, really really sucks. The guy didnt know anything about the card nor the options in the management software, and all he could do was say he would talk to engineering and see if they knew anything. Now though I am just posting this here now, I submitted a ticket last week with them, and their follow-up was non-existent.

3. Other World Computing was AWESOME (thats where I purchased the card). Their tech actually knew the card quite well, and knew a lot about RAID architectures. We figured out that the issue didnt happen in RAID 0,1, nor 3, just 5. It was stepping through some items with him that got us to the block size, and have us up and running right now. They were also ready to ship another card out just in case as well.

So for those considering rolling your own, its probaly going to be fine, but if its not, be ready to be left in the cold to figure it out. Luckily i have fiber here and this was a non-mission critical attempt. But there is something for having one person to call who would take responsibility for issues like this.

For those considering the 3522, i am at this point a bit concerned about trusting it. I will post back once i have had a chance to put it through the paces. Highpoint did say that these drives were now certified to be compatible (7200.11 1TB), but I would probably go with the Samsungs to do it all over again. The case and cables are certified, so this should have worked.

[Scott Thomas]

I have the RocketRaid 3522. I went with the iAge case and populated it with 8 500GB Seagate 7200.11 drives. I went with RAID 6 and let the block size at 64k. I’ve not had the problems you’ve described.

I thought about experimenting with the block size, but I never got back to that.

[Nils Fridén]

I run into exactly the same problem, but the weird thing is that it worked perfectly for about two weeks. After two weeks and 10 min, (were the last ten minutes was the time it took my client to give me final approval for some major online work.) I could no longer collect or copy my media. Sure I had backups, but it would had taken days of tedious work get everything back in order. Since I had nothing to lose, I used the migration function to change the raid from 128k to 64k block size without reformat. This took about 40 hours or so. Now my raid is working perfectly again and all media seems to have survived the migration. The info in your post saved me a major headache and about a week of work. Thank you very much!

[Randall Brown]

glad it helped you, but I would still keep up with your judicious backups, there is something wrong here. as now I am having even more issues. Until I can get it sorted I am using the Fibre stuff, but this is concerning.

My array worked fine as well for a couple of weeks before these issues. The switch to 64k worked for a while, but now the issues are back. I have now migrated to RAID 0 to see if that helps, but I am staring at a spinning beachball in final cut right now as I type this, so its not looking good.

A couple of questions for you.

1. What drives (Brand and size) and firmware are you using?
2. What RAID Level are you running?
3. RAID Block size, I see its now 64k before that it was 128k, did you ever go larger?

Their tech support is RIDICULOUS. Its really unfortunate, this seems like the killer card. I will update once I figure out this latest round of issues.

[Nils Fridén]

Thanks for the warning. I almost thought I was out of trouble after the migration.

I’m running 8 x 750GB Samung drives in RAID 5 (SAMSUNG HD753LJ, Rev. 1AA01108)

I tested the volume with 256k, 128k and 64k when setting up. After testing i settled for 128k and did a complete reformat in 128k. Worked great for about two weeks until last thursday when suddenly all file transfers or collect using FCP/Shake would freeze the finder or app after about 500MB or so. I havent changed or updated anything on my system during this period.

RocketRaid 3522 firmware: v1.2.13.25

[Randall Brown]

just a quick update. I got through to someone at highpoint who has been VERY helpful, but still no resolution, yet. I will post back once we have something final figured out.