May 28, 201313 yr I am currently dumping about 14TB of media back on to a fresh unRAID build. When I last dumped media it was transferring at an average of 40MB/s without using a cache drive. The box currently does not have a parity drive assigned as I want to dump the media, enable cache and then enable my parity drive for a fresh parity sync. However, I noticed today the transfer speeds have dipped into the 2MB/s arena and there is nothing different configuration wise or add-ons. It's running identical to how it was when I first started. There are no network changes either. Any off the tip of your brain ideas?
May 29, 201313 yr I think jonathanm hit the nail on the head. Odds are very good that your server has > 4GB of RAM and it's one of the boards that are impacted by this. Simply reboot your UnRAID server, and at the initial prompt select the option for 4GB boot (this limits the RAM to 4GB). It's VERY likely that will resolve your problem.
May 29, 201313 yr Assuming the 4GB option resolves this, just add the following to your SysLinux.cfg file: default menu.c32 menu title Lime Technology LLC prompt 0 timeout 50 label unRAID OS menu default kernel bzimage append mem=4095M initrd=bzroot label Memtest86+ kernel memtest
May 29, 201313 yr Any off the tip of your brain ideas? You've given symptoms, but provided no clues. Have you changed unRAID versions since you last "dumped data" to the array? Do you think letting us know the version of unRAID you are using, or attaching a syslog and the output of ifconfig eth0 would help us with some clues? (hint... they would)
May 30, 201313 yr Author Any off the tip of your brain ideas? You've given symptoms, but provided no clues. Have you changed unRAID versions since you last "dumped data" to the array? Do you think letting us know the version of unRAID you are using, or attaching a syslog and the output of ifconfig eth0 would help us with some clues? (hint... they would) Too much logic Joe. I was outside and no where near my unraid when I submitted the post. I still do not have my syslog available, but no I have not changed unraid versions. Also, the box currently only has 4GB of ram total.
May 30, 201313 yr The box currently does not have a parity drive assigned as I want to dump the media, enable cache and then enable my parity drive for a fresh parity sync. ... there is nothing different configuration wise or add-ons. It's running identical to how it was when I first started. There are no network changes either. So NO changes to the configuration; NO changes to the network; and NO change to the UnRAID version. And NO add-ons (or at least no changes to them). [i note you also said this in your initial post: "... there is nothing different configuration wise or add-ons. It's running identical to how it was when I first started. There are no network changes either." ] A few questions ... (a) You indicated you were getting 40+ MB/s and are now getting ~ 2 MB/s. Is this for the same basic type of transfers? i.e. were you transferring large video files, and are you now transferring thousands of relatively small files? (b) If you have UnMenu installed, have you run a SMART check on the drives? If not, install UnMenu and do so © The most "obvious" cause of this wasn't the case [updating UnRAID to one of the versions that's having issues with > 4GB of RAM] ... but another frequent cause is a missing pair in the cable (broken; disconnected; poor connector; etc.). If you've done anything that may have moved the cable, this could be the cause. Try a different Ethernet cable ... from the PC to your router AND from the UnRAID box to the router if possible. If you don't have another cable, at least unplug/replug the current cables, in case the connector got "jostled" and has a marginal connection to one of the pairs (using a different cable is a far better test).
May 30, 201313 yr Author version: 5.0-rc8a memory: 4GB total $ ifconfig: eth0 Link encap:Ethernet HWaddr 00:1d:92:b6:91:46 inet addr:10.0.1.50 Bcast:10.0.1.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:215119 errors:0 dropped:0 overruns:0 frame:0 TX packets:31555 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:320394296 (305.5 MiB) TX bytes:5302171 (5.0 MiB) Interrupt:45 Base address:0x4000 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:50 errors:0 dropped:0 overruns:0 frame:0 TX packets:50 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:2842 (2.7 KiB) TX bytes:2842 (2.7 KiB) $ smartctl -a -d sat /dev/sdX attached. They are named sdX.smart within zip. Syslog is attached Additional troubleshooting - prior to post(s): check switch check network cables, tested fine, replaced anyways check SAS8087 fanout cables, snug, seem fine test connection from desktop doing xfers to another box, transfer fine test connection to unRAID from another box, slow transfers as well perform memory test - memory fine after 3 days of testing disable all addons and test - still slow scratch head endlessly until scabs appeared, done. So, I know there is likely something I am missing, but I cannot figure out what it is. So, without further hesitation.. RIP INTO ME, what am I forgetting guys? smartctl_logs.zip syslog-20130529.zip
May 30, 201313 yr Author A few questions ... (a) You indicated you were getting 40+ MB/s and are now getting ~ 2 MB/s. Is this for the same basic type of transfers? i.e. were you transferring large video files, and are you now transferring thousands of relatively small files? - this is for the same type of files. (b) If you have UnMenu installed, have you run a SMART check on the drives? If not, install UnMenu and do so - I do have unmenu installed, but I still do smartctl from cli. © The most "obvious" cause of this wasn't the case [updating UnRAID to one of the versions that's having issues with > 4GB of RAM] ... but another frequent cause is a missing pair in the cable (broken; disconnected; poor connector; etc.). If you've done anything that may have moved the cable, this could be the cause. Try a different Ethernet cable ... from the PC to your router AND from the UnRAID box to the router if possible. If you don't have another cable, at least unplug/replug the current cables, in case the connector got "jostled" and has a marginal connection to one of the pairs (using a different cable is a far better test). - I have not upgraded unraid itself. I do not have more than 4GB of ram in this build. Cables are fine, replaced to test as well. See additional troubleshooting in previous step.
May 30, 201313 yr sdg has pending sectors and needs to be rebuilt. Check the wiki. Thee can be no pending sectors on any unRAID disk. After sdg is rebuilt make sure the pending sector count is zero. If you have a spare the use it and run pre-clear on sdg.
May 30, 201313 yr sdg has pending sectors and needs to be rebuilt. Check the wiki. Thee can be no pending sectors on any unRAID disk. After sdg is rebuilt make sure the pending sector count is zero. If you have a spare the use it and run pre-clear on sdg. However, this is not the cause of the slow transfers... It is an issue you'll need to correct at some point before you have to deal with a disk failure.
May 30, 201313 yr ... test connection from desktop doing xfers to another box, transfer fine => So cable between desktop and switch/router is fine ... test connection to unRAID from another box, slow transfers as well => So cable between switch/router and UnRAID could have an issue (did you try a different cable?); OR the issue is within the UnRAID server (probably more likely) I know you said "... check network cables, tested fine, replaced anyways ..." -> just wanted to confirm that was one of the cables you changed. The perplexing thing is that without a parity drive, UnRAID itself is practically out of the picture [Yes, it's managing the writes, but the core functionality of computing/updating parity and maintaining a fault-tolerant storage system isn't even in use !!]. Suggestion: I noted you're running an older RC ... save a copy of the flash drive (for easy reversion if you need to) ... and update it to RC12a and see if the symptoms are the same.
May 30, 201313 yr Author I am not using a parity yet temporarily while I copy the data over. Not too concerned with its integrity until its done as I have the source data on 1:1 backups. I'll upgrade and try. Going to replace my cache drive, I hate this 256GB SSD anyway, damn thing has like 10,000IOPS. and these wd 3tbs are shit in my opinion. If its about to fail that will be 2/2 in < 1 year
May 30, 201313 yr I am not using a parity yet temporarily while I copy the data over. Not too concerned with its integrity until its done as I have the source data on 1:1 backups. I'll upgrade and try. Going to replace my cache drive, I hate this 256GB SSD anyway, damn thing has like 10,000IOPS. and these wd 3tbs are shit in my opinion. If its about to fail that will be 2/2 in < 1 year Without a parity drive assigned, there's no reason to use a cache drive. Are you using WD Reds? I've had exactly the opposite experience -- I've got 6 of them and they've been VERY reliable ... good speed; no reallocated sectors; and a very nice warranty. Can't say the same for the Greens ... but I don't buy them anymore.
May 30, 201313 yr Author Okay, so the 2 WD 3tb I've had issues with are the cheap green drives. I have (5) 3TB reds sitting here that are in a cage, pre-cleared for this build as well as (4) 2TB red drives. All in all this build will be a touch over 30TB. I think I am going to copy what I can, pre-clear again!! Then upgrade CPU, Motherboard and memory. I've been wanting to put in this super micro board or tyan board sitting in my closet. Then, rebuild, reconfigure and copy stuff from a buddy who I set up unraid for a while back. It will be fine in the end, just frustrating. My primary unraid has been offline now for like 5 months. Everytime I get time to work on it and get things the way I want them, I find something breaking. The WD greens are coming out completely. Going to replace cache with either a sandisk ultra 128gb, mushkin chronos 240gb or samsung 840 pro 500gb. (I have a TON of hardware available, lol). Then, I should be fine moving forward.. we shall see. I am moving data between two windows boxes right now at 85MB/s, through the same switch as this unraid is on. One of the windows boxes is the source I was copying from as well, to unraid. I tested and replaced every network cable involved in this situation. I didn't see the pending sectors on sdg at all. Appreciate the words and advice. Will let you know how it turns out.
May 30, 201313 yr Sounds like you'll end up with an excellent system, albeit by following the path you had hoped to !! By the way, you need to change the title in your first post in the thread for it to show up on the list of topics ==> i.e. adding "Abandoned" to the last post doesn't make any difference in the listing under "General Support" ... if you change the first post's title, it will then show up as you intended
May 30, 201313 yr Author Gary: Yeah I realized that, but stepped away before I got back around so I could change it. I already have another unraid build running and running fine, but don't use it for media ;/ I think this is going to be a good time for me to go ahead and find a new case too, since I have a spare supermicro 5:3 cage + (5) more drives for it. I already ahve 15 drives in this one, going to pull the 2 wd 3tb greens I have left. The others in it are already cleared and just not in the array yet. So I will need 2 more drives to fill up 20 bays, but need a good full tower that has 12 5.25" bays. Right now I have an azza helios.
Archived
This topic is now archived and is closed to further replies.