January 23, 201115 yr unRAID Server Pro version: 4.5.6 (Upgraded to 4.7) 02/02/11 I am looking for some advise on what / how to establish why my music stutters / pauses when i am working with my other drives. Examples Access any other drive. - Music pauses for 3-6 seconds Use unrar after hdd is spun up. - Stutter for about 10 seconds Copy / transfer files. - Pause for 3-6 seconds Delete file through utorrent. - Pause 1-2 seconds and stutter for up to 10 seconds During transfers, copy, extraction or viewing files after the initial issue there is no other pausing or stuttering, it is only when I first start any process on the server. Build / setup details Server Mobo MSI P6N Diamond Ram 2 GB HDD's 22 I am using on board network card Cat 5a cable to my wireless router which is a d-link DIR-655 - info below I am using a windows XP sp3 pc to do all my work on with an asus p5w dh deluxe mother board. This also is connected via cat 5a cable. Also when i access anything in tower 8080 for example like switching page I get a short pause in my music. Nothing shows in my syslog that i can see as an issue which is attached. The most recent example would be when i clicked on additional options to add my syslog; I clicked the browse button and my music paused for 10 seconds. I just need some guidelines on where to start looking please. Thank you Settings for eth0: Supported ports: [ MII ] 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 auto-negotiation: Yes Speed: 1000Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: external Auto-negotiation: on Supports Wake-on: g Wake-on: d Link detected: yes NIC driver info (from ethtool -i) driver: forcedeth version: 0.64 firmware-version: bus-info: 0000:00:11.0 Ethernet config info (from ifconfig) eth0 Link encap:Ethernet HWaddr 00:13:d3:ee:b9:ee inet addr:192.168.0.102 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:284250592 errors:20 dropped:0 overruns:0 frame:20 TX packets:111290156 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:666420642 (635.5 MiB) TX bytes:2049467977 (1.9 GiB) Interrupt:33 Base address:0x4000 syslog-2011-01-23.txt
January 24, 201115 yr Author How are your spin-up groups defined? I have no spin up groups defined and I have my spin down set on all drives to spin down after 15 minutes of inactivity
January 24, 201115 yr What version of unRAID are you running? They've been there since 4.5 series. If you're running any recent version of unRAID, I suggest you look again. There are always spinup groups defined. Perhaps you're unaware of what they are. On unRAID 5.0 series, they're listed in each individual disk setting.
January 24, 201115 yr Author What version of unRAID are you running? They've been there since 4.5 series. If you're running any recent version of unRAID, I suggest you look again. There are always spinup groups defined. Perhaps you're unaware of what they are. On unRAID 5.0 series, they're listed in each individual disk setting. unRAID Server Pro version: 4.5.6 Default spin down delay: 15 minutes Force NCQ disabled: Yes Enable spinup groups: no Tunable (md_num_stripes): default 1280 Tunable (md_write_limit): default 768 Tunable (md_sync_window) 288 I set this up after i had some issues with my drives spinning down Joe.L pointed me in the right direction as I did not realise there was spin up groups set up by default.
January 24, 201115 yr What version of unRAID are you running? They've been there since 4.5 series. If you're running any recent version of unRAID, I suggest you look again. There are always spinup groups defined. Perhaps you're unaware of what they are. On unRAID 5.0 series, they're listed in each individual disk setting. unRAID Server Pro version: 4.5.6 Default spin down delay: 15 minutes Force NCQ disabled: Yes Enable spinup groups: no Tunable (md_num_stripes): default 1280 Tunable (md_write_limit): default 768 Tunable (md_sync_window) 288 I set this up after i had some issues with my drives spinning down Joe.L pointed me in the right direction as I did not realise there was spin up groups set up by default. Well, the reason you are seeing this is because of the lack of "spinup groups" being set to "Yes". Do a searching in the Announcements forum for "spinup groups" and you should find a discussion on the subject. It is an attempt to over some some of the Windows samba problems and a few others. I won't go into detail but essentially spinup groups will keep a disk spun up that is in the same group as any other so as not to produce IO conflicts on the bus while a drive is spun up (don't quote me if that is 100% correct... have not had my first coffee yet this morning).
January 24, 201115 yr What version of unRAID are you running? They've been there since 4.5 series. If you're running any recent version of unRAID, I suggest you look again. There are always spinup groups defined. Perhaps you're unaware of what they are. On unRAID 5.0 series, they're listed in each individual disk setting. unRAID Server Pro version: 4.5.6 Default spin down delay: 15 minutes Force NCQ disabled: Yes Enable spinup groups: no Tunable (md_num_stripes): default 1280 Tunable (md_write_limit): default 768 Tunable (md_sync_window) 288 I set this up after i had some issues with my drives spinning down Joe.L pointed me in the right direction as I did not realise there was spin up groups set up by default. Well, the reason you are seeing this is because of the lack of "spinup groups" being set to "Yes". Do a searching in the Announcements forum for "spinup groups" and you should find a discussion on the subject. It is an attempt to over some some of the Windows samba problems and a few others. I won't go into detail but essentially spinup groups will keep a disk spun up that is in the same group as any other so as not to produce IO conflicts on the bus while a drive is spun up (don't quote me if that is 100% correct... have not had my first coffee yet this morning). Correct... and the last time I gave advice, I think it was because you were complaining about all the disks spinning up. Joe L.
January 24, 201115 yr Yes, Joe L is correct. You have two base choices and one possible third option here: A) Live with the stuttering you're seeing now because of not using spinup groups B) Live with extra drives spun up to remove possible stuttering by using spinup groups C) Possibly tweak or rebuild your system so drives are not on the same controller channel As for the last option, how are the drives connected to the system? How are those SATA ports defined in BIOS? If they are connected to your motherboard SATA ports, make sure they are set to AHCI inside your BIOS. There might be a way to use SATA controller cards to provide different controller channels to minimize the number of drives on each spinup group.
February 2, 201115 yr Author Okay guys I have tried the following Put all my drive into varying spin up groups and had all drives spun up. No difference .. in fact when all drives are spun up the stuttering is continual when I am transferring files with Terra copy. Whereas if I have only the drives spun up I am working with the stutter / pause is only at the start of the copy process. Whether a drive is spun up or not when i access it I still get the 1-3 second pause when playing music. I have upgraded to 4.7 and tried the same again. No difference. I'm assuming it's a windows / network issue. Is there an easy way i can test this or a tool I can download to see / monitor what is going on. I would like to illuminate this as a network /windows issue before i investigate / consider a rebuild as suggested by BRit My bios is set to AHCI I am monitoring my power draw and I can see no issues there. Thank you
February 2, 201115 yr IMHO they are two possible solutions: 1. Disable the onboard LAN and use another Intel (or even Realtek based) NIC 2. If this does not help you may not like the second solution but I am going to at least suggest it. You are using an NVidia based motherboard and IMHO thou shall not use anything but strictly Intel or AMD based boards for their servers (especially full loaded) as you are going to experience one of the 'quirks' of these boards at one point or another.
February 2, 201115 yr Author IMHO they are two possible solutions: 1. Disable the onboard LAN and use another Intel (or even Realtek based) NIC 2. If this does not help you may not like the second solution but I am going to at least suggest it. You are using an NVidia based motherboard and IMHO thou shall not use anything but strictly Intel or AMD based boards for their servers (especially full loaded) as you are going to experience one of the 'quirks' of these boards at one point or another. Thanks bcbgboy What is interesting is when I copy from any drive in the array to my mp3 player attached to my pc USB port I get no stutter at all.. I am going to do a few more tests copying to and from my array to drives not attached to the server.
February 2, 201115 yr What is interesting is when I copy from any drive in the array to my mp3 player attached to my pc USB port I get no stutter at all.. I am going to do a few more tests copying to and from my array to drives not attached to the server. USB bus does not share anything with PCIe lane(s) AFAIK Forgot something - you have two onboard NICs - you have to disable any unused features on this premium board (ser. and par. ports, audio, firewire, IDE ports etc.). Then disable the secondary LAN too and try with only the first one. If does not work disable the primary and try with the secondary only as there are chances this one uses the PCI bus and this may help. If not disable both and try first with a PCI based gigabit NIC. If not working look here from your syslog for possible explanation why not to use NVidia boards: Jan 22 06:00:48 Tower kernel: forcedeth: Reverse Engineered nForce ethernet driver. Version 0.64. Jan 22 06:00:48 Tower kernel: ACPI: PCI Interrupt Link [LMAC] enabled at IRQ 20 Jan 22 06:00:48 Tower kernel: forcedeth 0000:00:11.0: PCI INT A -> Link[LMAC] -> GSI 20 (level, low) -> IRQ 20 Jan 22 06:00:48 Tower kernel: forcedeth 0000:00:11.0: setting latency timer to 64 Jan 22 06:00:48 Tower kernel: nv_probe: set workaround bit for reversed mac addr
Archived
This topic is now archived and is closed to further replies.