Announcement

Collapse
No announcement yet.

Barco icmp - hard drives upgrade

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Barco icmp - hard drives upgrade

    I want to upgrade my BARCO ICMP from 1TB drives to 2TB or more.... recommendations most welcome.

    or maybe link it to a BARCO DSL 100

  • #2
    For 2TB drives, please notice that only the Seagate ST2000NX0253 has been certified to work with the on-board controller chipset. I haven't checked if others have since been added, but that's what I get from the manual. Other drives may work, but you're on your own if you encounter issues.

    Having had quite some issues with Seagate drives over the past decade or so, I'm not that fond of them anymore, but I haven't heard anything bad about this particular range of drives yet.

    Comment


    • #3
      If you were to upgrade the RAID controller as well, you could use the 4TB SSDs that are mentioned here:
      http://www.film-tech.com/cgi-bin/ubb...003660/p3.html at the comment made 06-27-2019 07:30 AM Central
      You probably meant to write Dolby DSL100.

      Comment


      • #4
        That's a very expensive option, about 1000US$ per drive, + the upgraded RAID controller (any idea about the cost for that?)

        I guess I would upgrade to the 2TB seagate drives and add a cheap NAS for storage extension.

        Comment


        • #5
          Afaik the RAID upgrade is just a software upgrade.

          Comment


          • #6
            It's actually a new GEN 2 storage controller card (R9801854). A software update has to be performed with it.

            The weird thing is, the new GEN2 storage controller seems to be just a 3* SATA interface and is used as a software RAID, while the original ICMP storage controller was an actual hardware RAID card. It also carries the USB 3.0 ports of the ICMP(-X).
            The benefit of the new software RAID seems to be that it is more flexible with supported drive types.

            'There are two types of storage controllers with similar appearances:
            • Storage controller generation 1 (GEN 1), with a hardware RAID controller included. The old generation of ICMP is provided by default with this type of storage controller. New ICMP-X can't use this type of controller.
            • Storage controller generation 2 (GEN 2), without hardware RAID controller included (RAID is assumed by software). This type of storage controller support traditional hard drives (HDD) and solid state drives (SSD). Storage controller GEN 2 type is mandatory in ICMP-X, but it might find in old generation of ICMP after a hardware upgrade.'

            - Carsten
            Last edited by Carsten Kurz; 04-17-2021, 06:43 AM.

            Comment


            • #7
              Yeah, I've read up on it. It's a bit of an odd decision, but understandable somehow. The new controller actually is a "downgrade" to a straight SATA controller and the RAID is now handled in-software. While this may be a good deal for the newer ICMP with more processing power, I don't know if it's such a good deal for any older ICMP. Also, unless everything on your network is 10GE and the network is the primary way to ingest content, I don't see much added value for SSD right now.

              Comment


              • #8
                A lot depends on whether the ICMP in question is still in warranty. If it is, Cinionic have threatened in the past to be quite draconian, and to refuse any warranty claims if the log shows that a nonapproved drive model has been used, even if the reason for the warranty claim has nothing to do with the drives (though under California law, it's questionable whether such a refusal would be legal). If it's not under warranty, I can't see that you'd have anything to lose by simply trying any model of drive you like, and seeing if it works OK. I once put a set of four consumer desktop 4TB drives in an original (mk. I) ICMP: the venue needed 6TB of SMS storage for a festival, and that was the only option I could think of. Enterprise or NAS grade 2.5" 4TB drives simply weren't available. The consumer drives worked fine, and everything played without any glitches. The original set of Barco-approved 1TB drives were put back after the fest was done, though.

                Comment


                • #9
                  Much to think about. AND YES a typo BARCO DSL 100 .... I meant a DOLBY DSL 100. ICMP in question is outside any warranty

                  Comment


                  • #10
                    Typo: that should have been a set of three, not four, consumer 4TB drives in that Alchemy, obviously.

                    At the time of that festival, I scoured the Internet trying to find the cartridges that the drives sit in to slot into the Alchemy. Barco will not sell them empty, and I couldn't find them from any third party. So I had to fiddle about swapping the 1TB drives that came with the unit out of their cartridges and the 4TB consumer ones in, and then reversing the process at the end of their festival.

                    Comment


                    • #11
                      Yeah, imagine someone would come up with a bunch of industry standard 2.5" disk trays and hardware manufacturers actually using them... Ever since the invention of the standard 8" and 5.25" equipment bays by IBM, we never had a single standard for swapable drive trays...

                      The potential problem with consumer drives is that they may go into standby mode if not being written to or read from. This can cause havoc with a RAID. If the RAID needs to read or write to one or more drives in sleep mode, it needs to wait. Not only is this detrimental to the performance, the RAID could elect those drives for ejection from the RAID.

                      Besides that, in practice, I haven't seen much difference between "consumer" and "enterprise" drives in the last 25 or so years. The primary difference usually is just the warranty you get on a drive.
                      Last edited by Marcel Birgelen; 04-18-2021, 10:59 AM.

                      Comment


                      • #12
                        How enterprise drive handle errors is different. Enterprise drives, typically, won't re-read a bad sector, presuming that another drive has the data...and move on...improving their read speed. Consumer/desktop drives will re-read for the same data, presuming they are the only drive in the system.

                        I've seen huge longevity differences on consumer vs enterprise drives when they are on 24/7 and yes, it is reflected in their warranty.

                        When it comes to drives for cinema servers. read reliability is paramount to prevent "underflows." Underflows are what cause images/sound to stutter/freeze or otherwise be blemished. When a manufacturer "qualifies" a drive for their server, they are saying that it will work reliably in their server for the intended purpose. It isn't that other drives won't. However, they are not going to test every drive being made...once they have sufficient drives to handle their needs, those are the drives that have proven, at least during their testing, to meet the necessary criteria.

                        Here is one of the lists of approved Barco drives:

                        Screen Shot 2021-04-18 at 12.49.15 PM.png

                        As for Barco's drive caddy...yeah, Barco isn't so good about making lots of component parts available. They are more into assemblies. It wouldn't surprise me if they didn't have their caddies OEMed to them so the color matched and everything.

                        Comment


                        • #13
                          Originally posted by Steve Guttag View Post
                          How enterprise drive handle errors is different. Enterprise drives, typically, won't re-read a bad sector, presuming that another drive has the data...and move on...improving their read speed. Consumer/desktop drives will re-read for the same data, presuming they are the only drive in the system.
                          Well, that's more dependent on the controller than on the drive, as the controller defines the command completion time per NCQ queue. If you have a healthy array, you want to keep it short and fail early to avoid performance issues, but if your array is running on its last legs, it's best to try to get the data off there, as there is no more left to fall back on. Mind you: lots of controllers aren't that sophisticated and it's primarily software RAID that's catching up on this sort of thing. Why do I know this crap? I've done quite some hacking on the early implementation of ZFS on FreeBSD back in the day, as this was what our storage was based on, before we could pay some storage company big bucks to make it their problem (mostly). ZFS was just ported from Solaris and still had tons of little issues, among them a lot of performance issues. One of the issues was that a single failed drive in an array of like 60 disks could bring your entire array to almost a halt, because it kept re-reading from this one broken drive and waiting for data. I learned that it was a combination of controller, driver and lastly OS, not so much hard-drive firmware that caused this.

                          What I've heard from someone working at WD is that both Enterprise and consumer drives are often exactly the same hardware, at least when it comes down to plain SATA drives, the SAS drives aren't really sold to consumers. The primary differentiator between drives is the performance and this comes mostly down to RPMs and the amount of cache on the controller of the harddrive itself. Even assigning a "usage type" to the drive is mostly really marketing b.s., the only differentiator here is performance, not really reliability. There may be slight differences in firmware, but anno 2021, they seem to be mostly negligible, as the market for hard-drives is one with a foreseeable end-date, without much more room for billion dollar investments in research and new plants... Backblaze, who runs hundred-thousands of drives also reported that they didn't find any noticeable difference between the reliability of consumer and enterprise hard disks. Now, this report is from 2017, but I doubt anything significant has changed since then.
                          Last edited by Marcel Birgelen; 04-18-2021, 01:05 PM.

                          Comment


                          • #14
                            Oddly, WD was my go-to company for consumer stuff. To this date, WD "Black" drives have never failed me...either in 2.5" or 3.5". WD Blue, that's another story. The WD Blue 2.5" drives have been a bit of a disappointment, particularly some sizes.

                            Since WD bought out and took over Hitachi...they went from my most reliable to "varies from batch to batch." Seagate, which has, historically, ALWAYS let me down, seems to be doing better on their Enterprise drives. I see that Barco has gone to Seagate on their 2TB and 4TB drives

                            I've noted that Dolby, on their approved drives, have gone to Toshiba and less so to WD/Hitachi.

                            Me, personally, for SSD 2.5" laptop drives, I've gone to Samsung as nobody seems to have bad things to say about them. I have yet to have one give me grief and recently switched a desktop drive out for a Samsung SSD with nothing but improvement.

                            Comment


                            • #15
                              For most drives, 'Read Retry' can be configured using manufacturer tools. They only come into play anyway after the drive starts to degrade. That's why consumer drives usually work nicely (even mixed drive types) when new.

                              Not such a big deal for 3.5" drives. However, the portable market has more or less transitioned to small and medium sized SSDs, and 'real' higher capacity 2.5" drives become obsolete. Not so many offerings for spinning disk 2.5" drives left.
                              Last edited by Carsten Kurz; 04-19-2021, 07:27 AM.

                              Comment

                              Working...
                              X