Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dynamite Monkey]

Short and sweet:  The NAS takes forever to allow an initial access.  It can be 15 - 20 seconds of polling the drive before it responds.  Once accessed, the NAS is very fast - around 38 - 40MB/s. 

If it weren't causing issues, I could almost live with it but I have some automated utilities that push files out to the NAS and they're timing out and aborting the file copy processes.

Obviously I've got the power management settings set properly.  (Drives never hibernate.)  They're two Seagate 1.5TB 7200RPM drives configured in RAID1.

Any thoughts?  I'm guessing it's a firmware glitch but I haven't seen anyone else mention it yet.

[Jack Ryan]

Short and sweet:  The NAS takes forever to allow an initial access.  It can be 15 - 20 seconds of polling the drive before it responds.  Once accessed, the NAS is very fast - around 38 - 40MB/s. 

That's interesting as I have an identical initial access delay with all 320 and 325s configured for Raid 1. In my case the drives power down and the delay is the time taken for the drives to spin up to speed. What is particularly irritating is that the drives spin up sequentially meaning that the delay is often double what it needs to be (fix please?).

Are you sure your drives aren't nodding off?

Jack

[JavaLawyer]

The delays each of you are describing sound like HDDs spinning up. I observe an identical delay in my two DNS-343. You can validate whether the drives are sleeping or spinning by physically inspecting the HDDs as you access them.  Listen for audible queues and touch the drive before/during/after access and feel for vibration.

[Jack Ryan]

Yes, I stated that my delay was caused by the drives spinning up to speed.

What annoys me is that the (Raid 1) drives spin up to speed sequentially which, in my view, is a bug or at least, an over sight.

Jack

[JavaLawyer]

Yes, I stated that my delay was caused by the drives spinning up to speed.

I know...I was referring to the original poster (i.e. a confirmation of your observations)

What annoys me is that the (Raid 1) drives spin up to speed sequentially which, in my view, is a bug or at least, an over sight.

I don't think the DNS-343 has this issue. There are four HDDs, and from the observable delays, only the accessed HDD is spun up while the others remain dormant.

[JavaLawyer]

Gave your issue some more thought. My HDDs are formatted as four standard volumes rather than RAID, so there are no dependency between drives. Can't say if this issue would occur in my model if I were using RAID.

[Dynamite Monkey]

Thanks for the responses - I agree... it definitely seems like a spool-up delay.  The setting is definitely disabled though.

In case it's a corrupt setting, I enabled power management, set it for 10 minutes, set it for 5 minutes and then disabled it again.  I'll leave it overnight and will check it again tomorrow.

I'd love to physically check the drives but the NAS is located in an area which makes it impractical to do so.  If I still have this issue, I'll remove the NAS from its location and will be able to physically check it at that time.

Jack:  When you say you have the same issue with your 320s and 325s, do you have the power management enabled or disabled?  I'd very much like to get confirmation as to whether this is a firmware issue.

[JavaLawyer]

Glad I could provide some insight. Also, do you have firmware v1.00 or v.101? I'm not sure of the difference between the two versions as I have not seen the release notes.

[Dynamite Monkey]

My unit shipped with 1.01

I haven't tried the beta firmware yet.

I don't know whether enabling and disabling the power management has worked yet - the drive has been in fairly constant use tonight so wouldn't have spooled down / timed out etc. 

[JavaLawyer]

I don't know whether enabling and disabling the power management has worked yet - the drive has been in fairly constant use tonight so wouldn't have spooled down / timed out etc. 

If you see a difference, please post your results.

[lyian]

Humm.. Well not sure about everyone else, but i kind of expect this. My roommate has an external drive with power management, and i (currently) have a server that puts the drives to sleep after so long. In both cases, it takes time to spin them back up.. If this NAS does the same thing, it wouldn't really bother me that much so long as it wasnt too long. 15-20sec does seem a bit longer than im use to (the external usually takes about 5-10 seconds to spin up)

Java, what is the average delay of the drives on Your NAS?

[JavaLawyer]

Java, what is the average delay of the drives on Your NAS?

The HDDs in my DNS-343s are typically asleep. For first access, there is a 5-10 second lag, after which everything is instantaneous. Keep in mind that my 8 HDDs are formatted as Standard Volumes, so there are no interdependencies between HDDs. One of the discussion points here is whether the DNS-325 serially wakes HDDs or wakes/spins up in parallel. Also, if the DNS-325 spins up drives serially, does this only occur for RAID configurations where there are HDD interdependencies?

[Dynamite Monkey]

Finally had a chance to test it after an overnight lull.  It still has a long pause upon first access. 

I'm guessing it's a firmware bug at this point.  I'm going to get the drive out sometime soon so I can hear if the drives are spooling up during the delay. 

[JavaLawyer]

Again, from what you describe, it sounds like your NAS is spinning up, but you say hibernation is deactivated, which is throwing me for a loop. The only true way to find out if the HDDs are involved is a physical inspection of the device.

[JavaLawyer]

Finally had a chance to test it after an overnight lull.  It still has a long pause upon first access. 

I'm guessing it's a firmware bug at this point.  I'm going to get the drive out sometime soon so I can hear if the drives are spooling up during the delay. 

Did you have an opportunity to physically inspect the drives?