**** BEGIN LOGGING AT Mon Aug 01 23:59:56 2005 Aug 02 04:20:51 I've just requested logging for this channel at loglibrary. So if Gernika comes by asking, it's okay. Aug 02 05:13:37 anyone here tried to compile openslug lately? Aug 02 11:54:39 hi all Aug 02 12:20:53 hi Aug 02 12:21:17 I'm having problems with max size of files in unslung 5.5 Aug 02 12:21:26 is there a 2GB limitation? Aug 02 12:21:53 I can tranefer a 5GB file to the drive mounted in winxp Aug 02 12:22:14 the slug is then unable to access the file by ls or chmod Aug 02 12:22:45 if Iftp out of the slug to pc via ftp, the I see the following error: Aug 02 12:23:01 ftp> get bond20.VOB Aug 02 12:23:01 local: bond20.VOB remote: bond20.VOB Aug 02 12:23:01 200 Port command successful. Aug 02 12:23:01 150 Opening data connection for bond20.VOB (6117410816 bytes). Aug 02 12:23:01 bond20.VOB: short write Aug 02 12:23:01 426 Cannot retrieve. Failed. Connection reset (10054). Aug 02 12:23:04 2147479552 bytes received in 585 secs (3.6e+03 Kbytes/sec) Aug 02 12:23:17 this limit is 2047.99mb Aug 02 12:23:33 is there any way to fix this? Aug 02 12:49:24 Hi I cant get my slug to use the rootfs off the HD. Aug 02 12:49:33 I have unsling'd Aug 02 12:49:45 I have a .sda1root in / Aug 02 12:50:04 and a rootfs on the hardisk Aug 02 12:50:10 any ideas Aug 02 12:50:20 oh and i have followed readme Aug 02 12:51:15 version 5.5 Aug 02 12:57:14 oh well looks like I will have to narrow it down some more Aug 02 13:07:16 paulthego, you still here? Aug 02 13:08:43 paulthego, what happens when you power up the nslu2 with the hdd attached? Aug 02 13:09:27 yup i am here Aug 02 13:09:58 it mounts the drive and I have a disk1 shared Aug 02 13:10:18 but df indicates that rootfs is in flash Aug 02 13:10:20 can you paste your df output here? Aug 02 13:10:35 k its just rebooting Aug 02 13:11:01 the hdd is powered up before the nslu2 is powered up, right? Aug 02 13:11:12 Filesystem 1k-blocks Used Available Use% Mounted on Aug 02 13:11:12 rootfs 6528 5320 1208 81% / Aug 02 13:11:12 /dev/mtdblock4 6528 5320 1208 81% / Aug 02 13:11:12 /dev/sda1 240190052 44412 237705448 0% /share/hdd/data Aug 02 13:11:12 /dev/sda2 116661 4143 111314 4% /share/hdd/conf Aug 02 13:11:13 # Aug 02 13:11:19 yes its on Aug 02 13:11:27 before powering up Aug 02 13:11:59 its going wrong in linuxrc Aug 02 13:12:07 your df output seems to indicate that /sbin/unsling has not been accomplished... Aug 02 13:13:10 well before this bit executes echo Aug 02 13:13:10 /bin/mount -rt ext3 /dev/$prefroot /mnt Aug 02 13:13:10 if [ -L /mnt/sbin/init -o -x /mnt/sbin/init -o \ Aug 02 13:13:10 -L /mnt/bin/init -o -x /mnt/bin/init ] ; then Aug 02 13:13:10 /bin/echo "Root filesystem will be mounted from /dev/$prefroot ..." Aug 02 13:13:11 else Aug 02 13:13:13 /bin/umount /mnt Aug 02 13:13:15 /bin/echo "Can't find valid rootfs on /dev/$prefroot, using jffs2 ..." Aug 02 13:13:17 mounted= Aug 02 13:13:19 prefroot=jffs2 Aug 02 13:13:21 prefroot is sda1 Aug 02 13:13:50 then after that prefroot ends up as jffs2 Aug 02 13:14:44 have you modded linuxrc? Aug 02 13:15:19 i have added an echo of prefroot to a file in /opt Aug 02 13:15:28 to try and narrow it down Aug 02 13:15:57 thats how i know its in that section Aug 02 13:16:30 that df output is *exactly* what you get if you boot nslu2 without hdd and then hotplug the disk Aug 02 13:17:00 I wouldn't have messed with linuxrc.... Aug 02 13:17:02 I know but I am sure I have done the unsling ok Aug 02 13:17:30 I have only put an echo in Aug 02 13:17:38 and that was just now Aug 02 13:18:02 check to make sure that /dev.state is populated Aug 02 13:18:51 ok i look Aug 02 13:19:00 brb... Aug 02 13:19:37 I have an sda and an sda2 but no sda1 Aug 02 13:20:39 whoops i have sda1 Aug 02 13:20:57 there should be lots of stuff in /dev.state Aug 02 13:21:32 (not just sda1, or sda2) Aug 02 13:21:40 i am doing a ls of /dev there are lots of entries Aug 02 13:21:54 not worried about /dev Aug 02 13:22:00 just /dev.state Aug 02 13:23:25 how ? if i do ls /dev.state there is nothing Aug 02 13:23:36 sorry i am new :( Aug 02 13:23:45 cd /dev.state Aug 02 13:23:48 ls -al Aug 02 13:24:15 ok just the . and .. entry Aug 02 13:24:24 uggh Aug 02 13:24:32 k... follow my lead here... Aug 02 13:24:37 ok Aug 02 13:24:43 halt Aug 02 13:24:50 (that should power down your slug) Aug 02 13:24:54 unplug the hdd Aug 02 13:25:38 k slug down and drive unplugged Aug 02 13:25:41 power it back up without hdd attached... Aug 02 13:25:46 and brb.... :) Aug 02 13:26:28 k powering it up. BTW Thanks :) Aug 02 13:28:28 k it just beeped Aug 02 13:28:31 telnet in? Aug 02 13:29:01 yeah telnet in Aug 02 13:29:17 sorry... having to acclimate some fish to an aquarium.... Aug 02 13:29:34 takes a bit of time and every four minutes have to add more water.... :) Aug 02 13:29:36 :) Aug 02 13:29:54 ok i am in Aug 02 13:30:00 ok, now that you're telnetted in, rm /.sda1root Aug 02 13:30:17 k Aug 02 13:30:21 sync Aug 02 13:30:28 k Aug 02 13:30:33 reboot Aug 02 13:30:42 (still leaving hdd unattached...) Aug 02 13:30:48 yup :) Aug 02 13:31:51 when it beeps, telnet in Aug 02 13:31:59 and then let me know Aug 02 13:32:05 ok i will Aug 02 13:32:34 ok I am in Aug 02 13:32:46 ok, now do the cd /dev.state Aug 02 13:33:04 ok lots of stuff Aug 02 13:33:12 good deal... Aug 02 13:33:24 alright, now hotplug that hdd in Aug 02 13:33:59 k i have plugged it in Aug 02 13:34:13 check mount and df to make sure that it's recognized Aug 02 13:34:23 echo Aug 02 13:34:23 /bin/mount -rt ext3 /dev/$prefroot /mnt Aug 02 13:34:23 if [ -L /mnt/sbin/init -o -x /mnt/sbin/init -o \ Aug 02 13:34:23 -L /mnt/bin/init -o -x /mnt/bin/init ] ; then Aug 02 13:34:23 /bin/echo "Root filesystem will be mounted from /dev/$prefroot ..." Aug 02 13:34:24 else Aug 02 13:34:26 /bin/umount /mnt Aug 02 13:34:28 /bin/echo "Can't find valid rootfs on /dev/$prefroot, using jffs2 ..." Aug 02 13:34:30 mounted= Aug 02 13:34:32 prefroot=jffs2 Aug 02 13:34:34 sorry Aug 02 13:34:38 # df Aug 02 13:34:39 Filesystem 1k-blocks Used Available Use% Mounted on Aug 02 13:34:40 rootfs 6528 5328 1200 82% / Aug 02 13:34:47 wrong paste.. :) Aug 02 13:34:57 ok i can see it Aug 02 13:35:06 brb... fish... Aug 02 13:35:07 paste sill no good tho :) Aug 02 13:35:36 k Aug 02 13:36:05 mount is /dev/sda1 on /share/hdd/data type ext3 (rw) Aug 02 13:36:46 Filesystem 1k-blocks Used Available Use% Mounted on Aug 02 13:36:47 rootfs 6528 5328 1200 82% / Aug 02 13:37:03 d Aug 02 13:37:34 my df wont paste Aug 02 13:38:03 but /dev/sda1 is there Aug 02 13:38:34 good deal Aug 02 13:38:50 ok, now we're going to "unsling" again Aug 02 13:38:56 ok Aug 02 13:38:58 do /sbin/unsling disk1 Aug 02 13:39:37 finished unslinging? Aug 02 13:39:40 Copying the complete rootfs from / to /share/hdd/data. Aug 02 13:39:40 19213 blocks Aug 02 13:39:40 Preserving existing ipkg database on target disk. Aug 02 13:39:40 Linking /usr/bin/ipkg executable on target disk. Aug 02 13:39:40 Creating /.sda1root to direct switchbox to boot from /share/hdd/data. Aug 02 13:39:51 just finished Aug 02 13:39:55 sync Aug 02 13:40:27 passwd ? Aug 02 13:40:28 it synced? Aug 02 13:40:48 the password is going to be uNSLUng now... Aug 02 13:40:49 how can i tell sync just comes back to prompt Aug 02 13:40:58 that's all it's supposed to do Aug 02 13:41:00 :) Aug 02 13:41:08 type reboot and cross your fingers Aug 02 13:41:19 :) Aug 02 13:41:27 I let u kno in a mo Aug 02 13:41:31 see to fish Aug 02 13:41:47 I have another 2 minutes Aug 02 13:43:12 brb Aug 02 13:43:33 k i am in :) Aug 02 13:43:55 rootfs 6528 5324 1204 82% / Aug 02 13:43:58 :( Aug 02 13:45:18 also ls of dev.state is blank Aug 02 13:45:27 ooh Aug 02 13:45:54 that rootfs of 82% indicates something out of issue.... Aug 02 13:46:21 that /dev.state gets populated when the nslu2 boots off of flash alone (so no big deal) Aug 02 13:47:03 was this a binary download or a built unslung? Aug 02 13:47:10 yup I know that linuxrc thinks the prefroot is sda1 at the start Aug 02 13:47:17 binary dload Aug 02 13:47:30 you know what I am about to suggest right? Aug 02 13:47:39 V2.3R29-uNSLUng-5.5-beta Aug 02 13:47:58 what building my own? Aug 02 13:48:08 or a reflash :) Aug 02 13:48:08 no, just starting from scratch... Aug 02 13:48:26 reflash *first* to linksys r25 Aug 02 13:48:38 then reflash back to unslung 5.5 Aug 02 13:48:54 hmmm been trying for hours :) Aug 02 13:48:59 you can disconnect the hdd for all of the flashing... Aug 02 13:49:25 well thanks for all the help Byron :) Aug 02 13:49:36 and once you've booted up with unslung 5.5 without the hdd... do the unsling again... Aug 02 13:50:00 if that doesn't work, then there's something funky on the hdd forcing the rootfs back to flash Aug 02 13:50:05 well do Aug 02 13:50:10 will even Aug 02 13:50:11 and a format of the hdd would be in order Aug 02 13:50:18 :) brb Aug 02 13:50:19 fish Aug 02 13:50:24 I formatted it earlier tonight Aug 02 13:51:55 I reckon its /bin/mount -rt ext3 /dev/$prefroot /mnt Aug 02 13:52:55 anyhow thanks again for the help and hope your fish are fine Aug 02 13:53:08 bye :) Aug 02 13:57:51 good luck Aug 02 13:58:10 thanks cya Aug 02 15:03:52 test2 Aug 02 15:15:39 Hello again. So, I've not quite added a serial port, but I've found another failure mode of my nslu2/hdd/enclosure combination: Aug 02 15:16:51 Drive is turned on before NSLU2 (as expected). NSLU2 takes a long time to boot. After it's booted, it behaves exactly as if the hdd was plugged in AFTER boot. Aug 02 15:16:58 Enabling telnet I found this in the log: Aug 02 15:16:59 ! scan_scsis Aug 02 15:16:59 do read capacity!!! Aug 02 15:16:59 read capacity ok!!! Aug 02 15:16:59 The capacity is 320173056 Aug 02 15:16:59 *** 49 Aug 02 15:19:10 hrm Aug 02 15:21:40 Anyway: as I was saying. The disk checks out just fine, with all the utilities I can throw at it. The enclosure works (as far as I can tell), but my laptop sometimes has trouble booting Knoppix if attached at boot-time. So it looks like some enclosure/linux compatibility problem? Aug 02 15:24:14 Ok, since my irc client crashed trying to post the log, I've uploaded it here: http://www.beermex.com/@spc/nslu2-log.txt Aug 02 15:25:57 (laptop booting from CD - sorry - should have made that clear. Knoppix from CD fails to get very far sometimes if the enclosure is attached from the start, sometimes works fine, sometimes doesn't === seems to be same behaviour with NSLU2 too, is sorta my point). Any ideas? ALI5621 chipset in the enclosure Aug 02 15:28:05 <- looking at the log Aug 02 15:28:46 that hard drive needs to be checked out Aug 02 15:28:48 ByronT-Away: this channel is now logged at loglibrary too. Aug 02 15:28:54 FYI. Aug 02 15:29:10 dyoung, I figured that when you "tested" Aug 02 15:29:10 ByronT-Away - the disk checks out fine with all the utilities I can throw at it... what would you suggest? Aug 02 15:29:57 dyoung, his nslu2 throws these errors when hotplugging the device: SCSI disk error : host 1 channel 0 id 0 lun 0 return code = 70000 Aug 02 15:29:58 I/O error: dev 08:02, sector 0 Aug 02 15:30:21 what kind of bridging chipset? Aug 02 15:30:23 ByronT-Away ... no, that's not right at all. My point was that I was *not* hotplugging Aug 02 15:30:35 dyoung - ALI5621 (like I said.. just up there?) Aug 02 15:31:01 stripwax_: I dont usually read this channel; and I'm just here to setup logging, so i have not read the backlog Aug 02 15:31:25 dyoung - ok. was only 6 minutes ago tho Aug 02 15:31:48 be nice to dyoung... serial is his friend and occasionally I am too Aug 02 15:31:50 that doesnt mean I read it. Aug 02 15:32:23 see? ;) Aug 02 15:32:57 I'm sorry. Since I hadn't seen you join or leave this channel I guess I assumed you had a client open and logging messages.. Aug 02 15:33:13 Should I repost my message, if you haven't read it? Aug 02 15:33:39 some patience... a *wee* bit Aug 02 15:34:11 ByronT-Away - sure, just wondering if you or dyoung want a repost or not .. ? Aug 02 15:34:12 when *my* ProdSlug hdd started showing errors like that, I found that its hdd was on the verge of death Aug 02 15:35:16 ByronT-Away - the disk checks out fine with all the utilities I can throw at it ... what would you suggest? Aug 02 15:35:25 ( "ProdSlug" ?) Aug 02 15:35:40 code 70000 usually means your disk is going funny, or the controller isnt controlling. Aug 02 15:35:55 What kind of errors would a genesys chipset throw up Aug 02 15:36:56 We were talking about this yesterday (NSLU2 hanging on boot - I don't have a serial mod so can't provide more info) - the above log came about when I pulled the power on the NSLU2, replugged, and turned on. Maybe disk controller had remained in a funny state from the previous hang? Aug 02 15:37:34 looks sorta like this: Aug 02 15:37:44 usb-storage: Bulk data transfer result 0x0 Aug 02 15:37:45 usb-storage: Attempting to get CSW... Aug 02 15:37:45 usb-storage: usb_stor_bulk_transfer_buf: xfer 13 bytes Aug 02 15:37:45 usb-storage: Status code 0; transferred 13/13 Aug 02 15:37:45 usb-storage: -- transfer complete Aug 02 15:37:55 over and over and over and over..... Aug 02 15:38:58 dyoung - ok thanks. well, I'm not getting that. But I am getting "journal_bitmap" errors on /dev/sda2 which cause it to be remounted readonly. Disk runs just fine outside of the enclosure on a regular IDE bus Aug 02 15:39:12 lemme see if I can get a log of that one too Aug 02 15:39:48 you have run fsck.ext3? Aug 02 15:40:00 ByronT-Away - every morning, like clockwork ... Aug 02 15:40:26 on all partitions? Aug 02 15:40:59 on the ones that have been remounted readonly with journal_bitmap errors. which is (except one occasion) always /dev/sda2 Aug 02 15:41:14 run it on /dev/sda1 Aug 02 15:42:28 but that log shows errors on both sda1 and sda2 Aug 02 15:42:33 Should I run fsck.ext3 from NSLU2 or from a different PC with drive attached over IDE? I did the latter last night (on /dev/sda1 and /dev/sda2, in fact). Aug 02 15:42:54 ByronT-Away - that log doesn't show ANY journal_bitmap errors (does it?) Aug 02 15:43:14 since the nslu2 is the one doing the burping, then I would do it from the nslu2 Aug 02 15:45:39 ByronT-Away - so there's two issues - one, the disk hangs on boot; two, /dev/sda2 (usually) periodically gets remounted readonly after using the slug for some period of time. The log I posted is probably yet a third case which has only ever happened once, I think. Since, after the burp in that log, the slug booted successfully AND neither /dev/sda1 nor /dev/sda2 were read-only Aug 02 15:46:20 ByronT-Away - to fsck.ext3 /dev/sda1 from NSLU2, I should shutdown, remove disk, boot, and then reattach disk first, is that correct? Aug 02 15:46:27 right Aug 02 15:46:33 or umount Aug 02 15:46:56 nvm... shutdown and reboot without disk Aug 02 15:46:58 ByronT-Away - can I umount root? Aug 02 15:47:02 ok Aug 02 15:48:30 ProdSlug was my previous NSLU2 (lightning took vengence and killed ProdSlug for good measure) Aug 02 15:49:35 <- have to head home Aug 02 15:49:46 ok. thanks for help.. Aug 02 15:50:19 hrm.. . after hotplugging, disk seems fine. fscks clean Aug 02 15:50:37 you may just have to keep an eye on it... Aug 02 15:50:46 you have anything that can read SMART status Aug 02 15:50:49 Have been (for nearly 12 months now...) Aug 02 15:51:13 to see if SMART shows "impending failure" - which is what mine showed after a short while Aug 02 15:51:21 ByronT-Away - "the disk checks out with all utilities I can throw at it" .. manufacturer's low level sector checks, SMART, .. everything seems healthy Aug 02 15:51:34 funky power Aug 02 15:51:36 way healthier, even, than a more recently-purchase IBM drive Aug 02 15:51:37 :) Aug 02 15:51:51 k, have to really head home now Aug 02 15:52:00 Seriously believe it must be controller/linux incompatibility Aug 02 15:52:07 ByronT-Away ok thanks... Aug 02 15:52:55 anyone else around that might have ideas..? Aug 02 15:54:52 Ok, yesterday I recreated the journal on /dev/sda2 and it seems like /dev/sda2 hasn't gone strangely-read-only since then. So maybe that's one down. NSLU2 still often hangs on boot; only power cycling at power connector fixes Aug 02 15:55:24 Oooh, does Unslung 5.5 mount /dev/sda2 sync by default now? Aug 02 16:06:07 landslide- know anything about trying to diagnose hdd-enclosure compatibility problems? :-p Aug 02 23:51:05 hi **** ENDING LOGGING AT Tue Aug 02 23:59:56 2005