NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
NV+ slows after 15-20 minutes [SOLVED]
I am experiencing some performance issues with my NV+
It is used for storing drive images created using the Acronis Drive Image 2010 CD. Backups work fine, the issue arises when I come to do a restore. It works well for 20 minutes or so with the hard disk light flashing rapidly, then Acronis starts to show an error saying that the backup source has "poor media quality". After clicking through the error many times, (I assume for retries) sometimes it will run normally again and then have other periods of freezing.
I assumed it was perhaps an out of memory error so upgraded the memory to 512kB, which seems to have had a marginal improvement in performance, but no real change.
What have I done to try and diagnose the problem ?
Changed the network cables
Rebooted the switch (Netgear GS108T)
Changed the ports being used on the switch and connected the Readynas / laptop direct to the router
My NV+ is running Radiator 4.1.7, has a static IP address and has 4 x Seagate ST32000641AS 2TB drives fitted running X-Raid. All disks are in good health.
After struggling to restore a 21GB image yesterday I decided to drag out my old Synology DS101g NAS and try that. I copied the image file across the network using CIFS which took 37 minutes (about 9MB/s) and then did the restore from the Synology. That worked fine, it completed more rapidly than it would have done from the Readynas at full speed and I had just one media quality error. My conclusion from this is that the acronis file is fine, the network cabling is fine, it's the reading from the NV+ that's the problem
Unscientifically, it seems that the NV+ is just bogging down due to memory issues, but I don't know that, which is why am here asking for help.
More details of my setup follows
Static IP
Jumbo frames off (even though it still sends jumbo frames, changing the setting makes no difference to this problem)
1GB full duplex NIC
MTU 1500
CIFS, NFS, FTP, HTTP, Rsync protocols ticked
ReadyDLNA on
NASmonitor installed (shows minimal memory use with about 88% cache)
Full journalling disabled
Fast CIFS writes on
I know my firmware is a few versions out of date, but the more recent updates generally seem to be aimed at newer hardware / apple environments which aren't relevant for me.
I have a secondary issue with both my NV+ and Duo RND2000 in that I can no longer use simply insert a USB stick and copy to the NAS using the backup button. The job always starts and then part way through copying large files it seems to eject the usb and then say it is no longer available ??. I should do another thread for this, but include for information here.
Grateful for any thoughts ?
It is used for storing drive images created using the Acronis Drive Image 2010 CD. Backups work fine, the issue arises when I come to do a restore. It works well for 20 minutes or so with the hard disk light flashing rapidly, then Acronis starts to show an error saying that the backup source has "poor media quality". After clicking through the error many times, (I assume for retries) sometimes it will run normally again and then have other periods of freezing.
I assumed it was perhaps an out of memory error so upgraded the memory to 512kB, which seems to have had a marginal improvement in performance, but no real change.
What have I done to try and diagnose the problem ?
Changed the network cables
Rebooted the switch (Netgear GS108T)
Changed the ports being used on the switch and connected the Readynas / laptop direct to the router
My NV+ is running Radiator 4.1.7, has a static IP address and has 4 x Seagate ST32000641AS 2TB drives fitted running X-Raid. All disks are in good health.
After struggling to restore a 21GB image yesterday I decided to drag out my old Synology DS101g NAS and try that. I copied the image file across the network using CIFS which took 37 minutes (about 9MB/s) and then did the restore from the Synology. That worked fine, it completed more rapidly than it would have done from the Readynas at full speed and I had just one media quality error. My conclusion from this is that the acronis file is fine, the network cabling is fine, it's the reading from the NV+ that's the problem
Unscientifically, it seems that the NV+ is just bogging down due to memory issues, but I don't know that, which is why am here asking for help.
More details of my setup follows
Static IP
Jumbo frames off (even though it still sends jumbo frames, changing the setting makes no difference to this problem)
1GB full duplex NIC
MTU 1500
CIFS, NFS, FTP, HTTP, Rsync protocols ticked
ReadyDLNA on
NASmonitor installed (shows minimal memory use with about 88% cache)
Full journalling disabled
Fast CIFS writes on
I know my firmware is a few versions out of date, but the more recent updates generally seem to be aimed at newer hardware / apple environments which aren't relevant for me.
I have a secondary issue with both my NV+ and Duo RND2000 in that I can no longer use simply insert a USB stick and copy to the NAS using the backup button. The job always starts and then part way through copying large files it seems to eject the usb and then say it is no longer available ??. I should do another thread for this, but include for information here.
Grateful for any thoughts ?
30 Replies
Replies have been turned off for this discussion
- I think I would try upgrading the firmware. You can always downgrade again if there is a problem. I'd also check the OS fullness (requires ssh) or alternatively delete all logs from frontview (system and backup).
Maybe also check the SMART stats for the disks.
The NAS shouldn't be sending jumbo frames if the setting is off and the MTU set to 1500. So that is strange. - Have you considered updating the firmware?
Don't think this will solve your problem, but probably advisable.
You could also test all the disks using vendor tools, if doing this only remove and reinsert with Nas off.
EDIT: didn't see last post from StephenB, my post doesn't add anything. - Thanks guys, some responses to your thoughts
StephenB wrote: I think I would try upgrading the firmware. You can always downgrade again if there is a problem
I have been holding off upgrading the firmware as I thought 4.1.7 was stable and I didn't see much in the way of improvements for my type of use. I also didn't want to lose my logs (though I have now downloaded them for backup/examination). Certainly nothing that I have seen in the change logs to suggest that there are any fixes for this type of problem. I may still do so just to tryStephenB wrote:
The NAS shouldn't be sending jumbo frames if the setting is off and the MTU set to 1500. So that is strange.
I helped a guy a couple of years ago by running some wireshark tests to look at the packets on my setup to see if we could replicate his problem. http://www.readynas.com/forum/viewtopic.php?f=21&t=62414 . He was the brains, I was just the brawn. The conclusion was that the firmware sends large packets even when jumbo frames is off. :o I think you responded on that thread tooStephenB wrote: I'd also check the OS fullness (requires ssh) or alternatively delete all logs from frontview (system and backup).
Maybe also check the SMART stats for the disks.
I have rechecked the SMART status. All drives look very good with no reallocated sectors or any unusual entries. I also used phpsysinfo to look at how much of the system partition is being used. It's only 24% used with 464 MiB from a 1.94GiB partition. The /dev/hdc1/ partition is mounted at / as ext3
Looking at the information today I can see that the memory is now being reported as follows:
Kernel 7%
Buffers 13%
Cached 17%
Yesterday the cached memory was up at 88%, but I assume that was cached drive reads waiting to be transmitted.
Disk swap is 0% today and was 1% yesterday if I recall correctly.
Does that help, is there anything else I can be looking at ? - Have you asked acronis about this?
I appreciate that it seems to work almost error free from another source, but the "poor media quality " error seems to be fairly common with Acronis. - When I have a chance I might do my own wireshark trace with my nv+ v1.
I've done restores with Acronis TrueImage with no problems, but that was from my pro-6.
Maybe use NasTester (http://www.808.dk/?code-csharp-nas-performance) to test the read speeds of both the DS106j and the nv+ - I have been reading the Acronis forums as well, I agree that the poor media quality problem does arise with TIB. My understanding is that is often in relation to dvd backups. I have seen other Readynas users with the same issue, and like me they don't think there is any issue with the quality of the media, given that it is coming from a RAID array.
I have previously transferred the .tib file to USB or hard drive and restored partitions from that media without issue. I also used to be able to restore partitions from the NV+ without issue and using the same live CD so therefore it isn't possible that there is any change in the software environment being used for the restore. That tells me (I think) that the backup file is fine, it's just getting the data from the file to the Acronis live CD that I use to resore partitions. It was a development of that though process that led me to dig out the Synology NAS and try that out. The test was using the same network connections, same restore methodology, same .tib file, just from a different source (with the copy file being made from the NV+ original)
I have been running the tests with NAS performance tester, results as follows
Synology DS101g......400mb file......Write speed 12.24 MB/s.........Read speed..........17.95 MB/s
Synology DS101g......8000mb file.....Write speed 12.35 MB/s.........Read speed..........17.99 MB/s
Readynas NV+.........400mb file......Write speed 13.29 MB/s.........Read speed..........34.95 MB/s
Readynas NV+.........8000mb file.....Write speed 14.22 MB/s.........Read speed..........33.05 MB/s
When I was running the tests on the NV+ I also monitored memory usage. Cached memory was around 8% when idle, when the test file was being generated on the NV+ it rose to ~51%, and when the write test was being performed it rose to 88% meaning all available memory was being used (I think)
At the end of the final test there were three temporary swap partitions created, each of 249.88MiB, but with only 112kiB remaining used in each partition:
/dev/hdc2
/dev/hdg2
/dev/hdi2
I haven't tried updating the firmware yet, before I do any further thoughts or suggestions ? - I don't believe there is any issue with the tib file - the "poor media quality" warning is probably something to do with a response to a read that Acronis thinks is too slow.
What makes this odd is that the NV+ is 2x the speed of the old Synology. Did you look at the network stats saved on the NV+?
Of course if the NV+ is sending JF despite the MTU setting, that could result in fragmented packets reaching the PC. That might create some lost packets/retransmissions which could maybe explain the error??? StephenB wrote: I don't believe there is any issue with the tib file - the "poor media quality" warning is probably something to do with a response to a read that Acronis thinks is too slow.
I agree, and judging by the infrequent blinking of the access lights on the NV+ when this happens I believe that the NV+ has just stalled.StephenB wrote: What makes this odd is that the NV+ is 2x the speed of the old Synology. Did you look at the network stats saved on the NV+?
That's an interesting comment. I recognise that the old sparc equipment is getting on a bit now, but always thought it was good in it's time. I think the Synology is probably a year or two older, but only has a 250GB IDE drive fitted so expected it to be slower anyway.
I didn't look at the saved network stats, but will happily do so & post them back if you can tell me where to find them - are they one of the log files that can be downloaded ?StephenB wrote: Of course if the NV+ is sending JF despite the MTU setting, that could result in fragmented packets reaching the PC. That might create some lost packets/retransmissions which could maybe explain the error???
I looked at the lost packet / unrecovered packet information yesterday and reset the counters a few times. From memory on a typical failed transfer there was about 1000 resnt packets of which 250 were unrecovered. That seemed high to me, but not fatal ?
If you want to do your own work with wireshark, or even download the original captures from the other thread here then that should (based on godfreydanels analysis) show that the NV+ does send oversize packets. I stepped in before to try and help him because he thought the throughput rates were very low. I found the same, and upgraded memory/ switch to give myself the best chance I could for decent rates. Now I am finding that the connection is both slow and worse, now unreliable for large acronis restores. I have been messing about with some 5-8GB uncompressed image files for an Acer ZG5 with SSD and they have worked fine, taking ~20 minutes. When restoring a 20GB image to a Toshiba M10 with 250GB SATA I would guess it gets to 6-10GB and then starts to stall.
If you can let me know where to get the network stats I'll look at those and anything else that may be useful
Many thanks- On the NV+ vs Synology speeds, I was going simply by your read speeds (which is what is relevant to the Acronis restore). You got ~15 MB/s for the Synology, and ~30 MB/s for the NV+. Both had about the same write speeds.
The NV+ was certainly very competitive in its day - the ARM-based systems of course provide much better performance at that price point today.
The stats are on Network->interfaces (show errors button). That might be the same as the "lost packet" information you were talking about.
If you were using wired ethernet, of course there shouldn't be any lost packets. - Thanks again. yes, the lost packet information was from the network> interfaces tab, and it is wired ethernet so there are packets being lost where there should not be. That's also part of the reason why I reset the couters a couple of times, just to verify how mant new packets were being lost. This is typical of what I would see
Auto-negotiation 0
Bad packets 0
Disconnect 0
False carrier 0
Idle errors 0
Link failures 0
Receive errors 0
Symbol errors 0
VLAN tags 0
TCP Retransmits 1094
Unrecovered TCP Retransmits 230
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
