GarrettC Posted July 4, 2011 Share Posted July 4, 2011 Lastnight I received 2 new hitachi drives and installed into my unraid machine, while I'm there I upgrade to the latest beta. Copied specified files (twice) but the system wouldn't boot. formatted and copied all new beta files to the flash drive, now the system boots and functions somewhat.. It takes literally 10-15 seconds for each page to load in the web gui, every function you try to do takes forever. (ie; changing settings, adding drives, spinning up drives... anything) I did change the local master setting to YES but no change. Write speeds are less than 1MB/s if it even goes at all, sometimes it will stop all together at which point the web gui will time out. Read speeds are anywhere from 400KB/s to 80MB/s, its all over the place. I've tried reloading the os several times with the same result. Have also reverted back to the original setup (before adding drives) but running the updated beta, same result. I didnt make a backup of the previous OS so I can't get it back to what was working. I've attached the syslog but its huge (was bringing drives off/on line at the time sorry). System specs: Asus P5KPL-CM Intel E7400 cpu OCZ 4gb DDR-400 Intel gig ethernet pci-e (2) WD2001FASS 2tb (2) Samsung HD204 2tb (2) Hitachi HDS 2tb Cat6 to gigabit switches syslog_unraid5.0b7.txt Link to comment
Joe L. Posted July 4, 2011 Share Posted July 4, 2011 sounds like you have a LAN issue, especially if it takes a long time for the web GUI to load in the browser. Link to comment
GarrettC Posted July 4, 2011 Author Share Posted July 4, 2011 Whats weird is that everything was great up until the beta7 install. Write speeds were 30+, Read speeds were 80-120. This was without a cache drive also... im thinking about going back to an older version just to see whats going on. Link to comment
Joe L. Posted July 4, 2011 Share Posted July 4, 2011 Whats weird is that everything was great up until the beta7 install. Write speeds were 30+, Read speeds were 80-120. This was without a cache drive also... im thinking about going back to an older version just to see whats going on. did you switch the flash drive? What do you see when you type: ethtool eth0 Link to comment
GarrettC Posted July 5, 2011 Author Share Posted July 5, 2011 Same flash drive as before, just loaded the 2 upgrade files then eventually reloaded the whole thing. Settings for eth0: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised pause frame use: No Advertised auto-negotiation: Yes Speed: 1000Mb/s Duplex: Full Port: Twisted Pair PHYAD: 1 Transceiver: internal Auto-negotiation: on MDI-X: on Supports Wake-on: pumbag Wake-on: g Current message level: 0x00000001 (1) Link detected: yes Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 What's up with all the "unRAID driver removed" stuff in the syslog? That doesn't look familiar. Joe, I can't help but wonder if the the new kernel and drivers from the latest beta is having an effect... Oh, can we get ethtool -S eth0 or ifconfig eth0? Those will show stats. Link to comment
Joe L. Posted July 5, 2011 Share Posted July 5, 2011 What's up with all the "unRAID driver removed" stuff in the syslog? That doesn't look familiar. Joe, I can't help but wonder if the the new kernel and drivers from the latest beta is having an effect... It could, but nobody else has reported something similar. From what I understand, the new "md" driver loads and un-loads itself in order to support warm-swap hard-drive installation. You can swap out drives while the array is stopped, without powering down, if your hardware supports it. Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 Ah, okay. About the kernel/drivers, having yet to figure out any of these performance things this weekend I'm getting ahead of myself. Link to comment
GarrettC Posted July 5, 2011 Author Share Posted July 5, 2011 What's up with all the "unRAID driver removed" stuff in the syslog? That doesn't look familiar. Joe, I can't help but wonder if the the new kernel and drivers from the latest beta is having an effect... Oh, can we get ethtool -S eth0 or ifconfig eth0? Those will show stats. NIC statistics: rx_packets: 10787 tx_packets: 517 rx_bytes: 967918 tx_bytes: 51600 rx_broadcast: 8640 tx_broadcast: 42 rx_multicast: 1627 tx_multicast: 0 rx_errors: 0 tx_errors: 0 tx_dropped: 0 multicast: 1627 collisions: 0 rx_length_errors: 0 rx_over_errors: 0 rx_crc_errors: 0 rx_frame_errors: 0 rx_no_buffer_count: 0 rx_missed_errors: 0 tx_aborted_errors: 0 tx_carrier_errors: 0 tx_fifo_errors: 0 tx_heartbeat_errors: 0 tx_window_errors: 0 tx_abort_late_coll: 0 tx_deferred_ok: 0 tx_single_coll_ok: 0 tx_multi_coll_ok: 0 tx_timeout_count: 0 tx_restart_queue: 0 rx_long_length_errors: 0 rx_short_length_errors: 0 rx_align_errors: 0 tx_tcp_seg_good: 0 tx_tcp_seg_failed: 0 rx_flow_control_xon: 0 rx_flow_control_xoff: 0 tx_flow_control_xon: 0 tx_flow_control_xoff: 0 rx_long_byte_count: 967918 rx_csum_offload_good: 10727 rx_csum_offload_errors: 0 rx_header_split: 0 alloc_rx_buff_failed: 0 tx_smbus: 0 rx_smbus: 0 dropped_smbus: 0 rx_dma_failed: 0 tx_dma_failed: 0 Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 Looks good. Obviously something is broken but how are you testing performance? Have you tested local transfers for comparison, either at the console or via telnet? Try hdparm -t /dev/sdX for reads of each drive. Is the system doing something you aren't aware of? Try uptime, specifically the load averages. Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 It's kind of tough to bounce between 4th barbecue and troubleshooting. I can't help but go back to your initial report that it all started with the two new drives (and b7). I've had drives die in ways that took down the host interface and caused horrible/weird performance for the whole system. You've probably already tried removing those new drives, right? Link to comment
GarrettC Posted July 5, 2011 Author Share Posted July 5, 2011 lol yea we just finished bbq'ing and blowing up a years salary in fireworks (seems like it ) I removed the controller card and 2 new drives but still suffered terrible performance. I'm gauging the speed on how the gui loads and the generic report as windows is moving the files. Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 Are both unRAID and your Windows machine on the same switch? Have you tried a different switch, or at least different ports? Other client systems? You might be comfortable with that part of your setup but it still feels like a network problem from here. Can you do the hdparm tests "because it's easy"? Log in at the console and: htparm -t /dev/sdX Do it for every connected drive. I saw abcefgh before you removed the two. Link to comment
GarrettC Posted July 5, 2011 Author Share Posted July 5, 2011 Typing up a reply with all that I have found... oh yea.. your gonna love this LOL. Cliffnotes = bad drive(s) but not the new ones First I downgrade to 4.7 and BAM all the problems are gone... for a few minutes then the speeds start to act up again. So I telnet in and check the drive stats, MD2 has errored and so on. I run down to check it out and can smell HD Death, CRAP! Turns out issue #1 was a failing cache drive, temps didnt reflect it but she was in thermal meltdown! I think "Great" problems solved right? NOPE, im still on 4.7 at this point so I think maybe the files have been corrupted.. I load up 5.0b6 and things are great for a few minutes then go to crap again. WTF at this point im ready to light it up like the fireworks we'd been blowing up all night!!! So, I load up a fresh batch of 5.0b7 which for some reason completely pissed off the network and locked up the modem, oh yea we're havin fun now! I get everything reset and load the webui. Its fresh and no drives assigned at this point and baby SHE'S FLYIN. I begin testing each drive one by one to see which one slows the system down. I hit drive #4 (WD 2tb 2mos old) and it slows to a halt. I'm able to load all drives but that one and the box is better than its ever been. I'm seeing write speeds of 80MB/s and read speeds of over 120MB/s, this is without a cache drive too. Never seen a system move this fast. I guess it didnt like being messed with the first time around, when the drives came back up it was not a happy camper. Now I need to see if the WD is bad or if its just got a few misplaced files with errors? Link to comment
cyrnel Posted July 5, 2011 Share Posted July 5, 2011 That sounds like a painful 4th. Glad you figured it/them/everything out. Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.