pkelecy

Member
  • Content Count

    23
  • Joined

  • Last visited

Community Reputation

0 Neutral

About pkelecy

  • Rank
    Member
  1. I'm have a Xeon workstation that would like to re-purpose as a NAS server. I am planning to put new SSDs in it, rather than HDDs, as I don't have that much data, and I would prefer the reliability and low noise of an SSD (it will be sitting 3ft away!). My question is whether I would likely encounter any issues using a data center SSD (like a Samsung 883 DCT) rather than a consumer SSD. The data center SSD's are not that much more expensive, but have better MTBF and TBW specs. I ask because I've read not all SSD's (or HDD's) will work in all systems. I'm not sure why that is, and it's not something the spec sheets appear to mention. But I'd prefer not to find out the hard way! Thanks for any feedback on this.
  2. Just out of curiosity, I inquired with Linksys about the difference between these two products (given how very similar their design and specs are). This was their reply: ******************************************* Thank you for contacting Linksys Customer Support. With regards to that, I could say that the ability of the device can be seen on its actual peformance and not on its features. Thank you and have a nice day! ******************************************* Not a very satisfying response (basically "I don't know") but what else could they say.
  3. pkelecy

    Are two platters faster than one?

    Thanks for the feedback and links. This is an interesting topic, and it sounds like there are a number of possible avenues open for improvement. The real challenge though is to make improvements that don't (or minimally) increase the cost. No doubt it's tough to do. I suppose the first approach (which is what they do) is to improve the performance of the existing design: better modulate (or increase) rotational speed and improve the stroke control to reduce seek times. I suppose adding multiple heads (to the same control arm) would help also, but with more cost and complexity. To read from multiple platters at the same time would require being able to make fine adjustments independently, if I understood this correctly. This (fine adjustments only) could probably be designed into the current actuator mechanism more cheaply than adding multiple actuators, but still with a cost consequence. No doubt these ideas have all been considered (and discarded) by the drive manufacturaturers, but it's fun to think about anyway. Thanks again!
  4. Hi -This is something I was wondering about. Would a HD with two platters be faster than a HD with one (assuming the same areal density) because it can (in priciple) read and write both platters simultaneously? Also, along the same lines, is data access from tracks located near the outer edge of a platter faster than those located near the inner because of the faster (linear) head speeds at outer positions? Just curious, but if either of these are true, then it does imply that using multi-platter HD's with more capacity than needed would be a way to improve performance. Thanks.
  5. Thanks for the feedback. I think the Linksys Integrated KVM is a new product, but it's specs (and design) are so similar to the Iogear that I wonder if it isn't the same thing just repackaged. In terms of cost, Iogear currently has a $10 rebate on the GCS62, puting it at about the same price as the Linksys. But, the Iogear has a 3-year warranty, vs 1-year for the Linksys product. So Iogear seems the way to go. Thanks again!
  6. Hi, I'm trying to decide between these two 2-port KVM switches. Both have built-in cables and very similar specs. What Im not sure about is whether the Linksys will work with wheel mice (I know some other Linksys models didn't) and whether the Iomega will retain settings (caps, num, scoll) when switching between PC's. Also, I've heard some KVM's have trouble doing stuff like allowing one PC to reboot while switched over to the other. I don't know if that's a problem with these or not. If anyone has used either of these I would appreciate hearing your opinions and if you know the answers to any of these questions. Thanks! -Pat
  7. pkelecy

    3ware 7000-2 vs 7500-4?

    Thanks Jan for the feedback. Based on what you've said, it sounds like the main benefit of the 7500-4 would be for future expansion, not performance. At this point, I don't expect to need more HD space for next year or two. Given how quickly the technology's advancing, once I outgrow this system I will most likely opt to just replace it. So it sounds like the 7000-2 is the way to go. Thanks again! Pat
  8. What about extra read/write heads per side - would'nt that speed things up also?
  9. In a simple 2-disk, RAID 1 setup (with two, 120GB IBM HD's), would the 7500-4 (with 64 bit PCI) perform better than the 7000-2 (with 32 bit PCI) because of its 64-bit bus? Trying to decide which to get (my server has both 32 and 64 bit PCI slots available). But it depends on where the bottle necks are. Any opinions? I emailed 3ware about it and weren't sure (so they said). Thanks. -Pat
  10. Sorry for the newbie question - something I'm not quite clear on. Do DSL routers require a separate DSL modem (located between the router and phone line) or is this normally built in to the router (i.e. phone line plugs directly into the router)? DSL just became available in my area and I would like to use a PPPoE router, but am not sure if I also need a separate DSL modem. I hope not, but some info I've seen indicates yes, others no. Thanks for the help!
  11. Hi, DSL just became available in my area, and I am looking for a good DSL router/modem which supports PPPoE (for use with a small server). Any recommendations? Also, any recommendations for other forums where I might post this? Thanks for the help! Pat
  12. Hi, Does anyone know of any good reviews comparing ATA RAID and SCSI RAID performance? I'm interested in knowing what the real performance differences are between these with current, SOTA technology (for both). I've only come across one review and it compared current (at the time) ATA RAID with older SCSI RAID. Obviously not a particularly valid comparison if trying to gage current capabilities. Thanks for any info. I appreciate it. Pat
  13. pkelecy

    How noisy are WD1200JBs or BBs?

    Thanks for all the feedback! It really helped and was just what I was looking for. I guess I'll place the order. Pat
  14. Hi, I'm looking for some perspective on a concern I have. I'm considering buying a Dell server (for a home office) with two 120GB ATA drives (RAID 1). Because I'll be sitting about 4 feet away from it, I would like it to be reasonably quiet. I know Dell uses WD, Seagate, and Maxtor drives (finding out which specific model it might be has be nearly impossible - no one at Dell seems to know). But based on reviews I've read, I think "worst case" would be WDs (most likely 1200JB's or 1200BBs). However, I don't have a good feel for how noisy a pair of WD's would be, and whether this would turn out to be minor compared to fan noise and other. Any opinions or suggestions? Thanks. I appreciate the help! -Pat
  15. After reading the review "IDE Raid Roundup" ( http://www.tech-report.com/reviews/2002q4/ideraid/) and the reported poor RAID 1 write performance of the 3ware 7500-4, I email 3ware about this. I thought their response might be of interest: "The poor write performance with the writes on RAID 1 arrays does not happen on Iometer. It does show up on the ATTO benchmark. We have seen this before. It is because the ATTO benchmark asserts the FUA bit, which 3ware acknowledges by sending flush commands to the disk drives after each write. This has the effect of slowing down 3ware write performance. In my experience the competition does not honor the FUA bit, so they don't show the same slow performance. We show top performance with RAID 1 writes under Iometer because Iometer is a better benchmark, and does not assert the FUA bit. Real life applications do not typically assert the FUA bit, so real life performance would more closely match the Iometer results." Any opinions?