Since then I have had extremely poor write performance from Win7 clients. Even when attempting to initiate a transfer to a share the machine can sit there waiting for ages before it commences copying, and then I get very poor write speeds in the order of 7-9 MB/sec (according to Windows 7 UI).
Doing a read off the NAS (copying to the clients) I can get 120-130MB/sec sustained no issues.
Copying to/from a Synology DS710+ sitting on the same switch as the Ultra 6 shows no performance changes to it (i.e. its not network hardware at fault). I can only ID the firmware upgrade as the source of the problem?
I have JF enabled and all ports to my managed switch (an HP Procure 1810G) are showing MT of 9216, and its error logs show there are not network errors happening.
Doing a direct connection yields ~120MB/sec read from NAS (using Win7 copy dialogue) and ~12MB/sec write to NAS (same file). Results are virtually the same as going via the switches.
I am at a loss to explain why I have excellent read speeds but rubbish write speeds all of a sudden.
You might want to contact NetGear Technical Support: http://www.readynas.com/support, and edit the title of the thread (i.e. subject of the first post in the thread) to include your case number.
Edit: Have you checked to make sure the NIC drivers for your Windows 7 PCs are up-to-date?
You might want to contact NetGear Technical Support: http://www.readynas.com/support, and edit the title of the thread (i.e. subject of the first post in the thread) to include your case number.
Edit: Have you checked to make sure the NIC drivers for your Windows 7 PCs are up-to-date?
Yep I've checked. Its across 3 different systems (Dell laptop, 1155-based desktop and a 1156-based desktop) all seeing the same low speeds.
Now that I think about it, this may have not coincided with the firmware upgrade, but I'm at a loss to explain what else would have caused this.
There is something wrong in the network connections between Win7 and the latest sets of ReadyNAS FW. I still struggle with this after a year or so. Sometimes the problem is the NIC driver, sometimes it is just restarting Win7 that solves the problem.
The symptoms are the same - a data transfer is initiated but nothing happens. Then after a while either it all starts or something times out and "Network Name is Invalid" (or something like that) is displayed. Normally the problem is on the NAS Write side so copying from NAS to Win7 is OK, but vice-versa fails.
I've reinstalled the NAS FW, changed drives, reinstalled Win7, updated drivers etc etc but the problem persists erratically.
The delay in starting the transfer is something to do with name resolution - if I type the IP address as the destination, it fires right away.
Still god-awful slow though. More updating and tweaking today can't see me past 12-14MB/sec transfers to the NAS; but I can copy stuff from it at awesome speeds.
If you download your logs, what's the oldest RAIDiator version mentioned in initrd.log?
[2011/02/22 11:42:31] Factory default initiated by Frontview! [2011/02/22 11:46:56] Updated from RAIDiator to 4.2.15. [2011/08/02 00:05:30] Updated from RAIDiator 4.2.15 to 4.2.17. [2011/08/05 01:22:35] Updated from RAIDiator 4.2.17 to 4.2.17.
I reapplied the firmware today.
And on reflection, I am pretty confident the poor write speeds (to the NAS) started after updating to 4.2.17 from 4.2.15. I just checked timestamps and a large 3GB file I recently transferred was before the update to 4.2.15.
WARNING: If you update to 4.2.17, you will NOT be able to go back to 4.2.15 or earlier due to the automatic change in the partition format to support 3TB+ hard disks.
I'd suggest you open a tech support case as I mentioned in an earlier post in this thread.
Tech support should be able to help diagnose the issue and suggest a solution.
WARNING: If you update to 4.2.17, you will NOT be able to go back to 4.2.15 or earlier due to the automatic change in the partition format to support 3TB+ hard disks.
I'd suggest you open a tech support case as I mentioned in an earlier post in this thread.
Tech support should be able to help diagnose the issue and suggest a solution.
Yes I read that after I posted the question. Bugger.
I'm was trying to avoid a support case. They were not that helpful the last time I called about something on my NV+. Hoping someone else knows the issue, or it is a known fault with .17 firmware.
The correct procedure is to work with tech support and then if you're getting nowhere with tech support to ask one of the Jedis for assistance.
Users like me can help out on the forums with the limited knowledge we have.
I wonder whether there was an issue with the GPT upgrade on your system? The GPT upgrade was thoroughly tested but that doesn't mean something didn't go wrong.
Disk /dev/sdd: 2930277168 sectors, 1.4 TiB Logical sector size: 512 bytes Disk identifier (GUID): XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX Partition table holds up to 128 entries First usable sector is 34, last usable sector is 2930277134 Partitions will be aligned on 8-sector boundaries Total free space is 5108 sectors (2.5 MiB)
Number Start (sector) End (sector) Size Code Name 1 64 8388671 4.0 GiB FD00 Linux RAID 2 8388672 9437247 512.0 MiB FD00 Linux RAID 5 9437256 2930272064 1.4 TiB FD00 Linux RAID
Disk /dev/sdd: 182401 cylinders, 255 heads, 63 sectors/track Units = cylinders of 8257536 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System /dev/sdd1 0+ 181688- 181689- 1465138583+ ee EFI GPT /dev/sdd2 0 - 0 0 0 Empty /dev/sdd3 0 - 0 0 0 Empty /dev/sdd4 0 - 0 0 0 Empty
Disk /dev/sda: 3907029168 sectors, 1.8 TiB Logical sector size: 512 bytes Disk identifier (GUID): C5BA6C32-8822-43FA-B5CB-1E0B6BC0C376 Partition table holds up to 128 entries First usable sector is 34, last usable sector is 3907029134 Partitions will be aligned on 8-sector boundaries Total free space is 5108 sectors (2.5 MiB)
Number Start (sector) End (sector) Size Code Name 1 64 8388671 4.0 GiB FD00 Linux RAID 2 8388672 9437247 512.0 MiB FD00 Linux RAID 5 9437256 3907024064 1.8 TiB FD00 Linux RAID
Disk /dev/sda: 243201 cylinders, 255 heads, 63 sectors/track Units = cylinders of 8257536 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System /dev/sda1 0+ 242251- 242252- 1953514583+ ee EFI GPT /dev/sda2 0 - 0 0 0 Empty /dev/sda3 0 - 0 0 0 Empty /dev/sda4 0 - 0 0 0 Empty
I probably will log a ticket early next week, do we get bumped to a call centre here in Oz?
I think that looks fine (you're going to have some higher numbers than I do as you have a 2TB disk).
Edit: If you have a ProSupport contract you can call them. Otherwise you still can but they'll charge you. Online Cases are free. See http://www.readynas.com/support
I haven't rung tech support ever so I don't know where they are based.
Your comment about Name Resolution set me to thinking about WINS. I set one of the NAS units to be a WINS server, set all other units to see it and rebooted. Seems to have solved the problem for the moment. I'm running 4.2.18 on an Ultra 6+ and 2x Pro Pioneers.
Okay, I got really frustrated with the transfer speed.
I'm sure I did this the other night, but I just turned off Jumbo Frames on the Ultra but left it on on the host machine (and left Ultra6 MTU at 1500).
Write speeds to the NAS are almost back to how they used to be.
I do not understand.
How can with JF enabled, read speeds off the NAS be excellent but writes terrible. With JF disabled, read speeds are good (not quite as good as with JF enabled) and writes are far higher - almost what they used to be.
Something screwy going on with this firmware, I am sure.
I note a couple of other threads along similar themes.
Look at your PC NIC settings, there's some mismatch. That's why Jumbo Frames is causing a problem. What's the NIC ? Also look at Flow Control - enabled or not ?
What's the complete Network config ? What switch(es) are in the link ? I have 3x Netgear GS108tV1s in the link - they need special settings for Port controls.
Look at your PC NIC settings, there's some mismatch. That's why Jumbo Frames is causing a problem. What's the NIC ? Also look at Flow Control - enabled or not ?
What's the complete Network config ? What switch(es) are in the link ? I have 3x Netgear GS108tV1s in the link - they need special settings for Port controls.
All the PC NICs have got JF enabled, flow control/green ethernet disabled.
I have got 2x RTL8111E motherboards (GA-H67N and GA-H55N motherboards), Realtek 8168B (Intel D510MO) and a Broadcom Netlink in my Dell laptop.
All 4 exhibit the same problems.
My central switch is a HP Procure 1810G (J9450A) which is a managed 24 port GigE switch.
The Ultra 6 (and a DS710+ Synology) hang off a Cisco SG 100D-08 8 port GigE switch which is off the HP Procurve.
There's other stuff in my network, but that's the link between these machines and the Ultra 6.
I get the same performance even when directly connected though (see earlier) - JF enabled at both ends, it crawls on the write to the NAS.
I had a mobo with the same Realtek NIC and a H55 chipset (Gigabyte Mitx) and I found that enabling Flow Control made an improvement. I also remember that there was an issue with the updated Win7 drivers. There were two versions - the one that Windows wants to install and the latest ones from Realtek. As the mobo has been replaced I can't replicate the answer, but try some alternatives.
I presume you can try these without the switches in the link as no doubt they will also interact with each other.
