- main - earth.li

udev and hdparm
by Barry Samuels 28 Jan '13

28 Jan '13

I have two sata backup drives which I use for daily rsync backups. As the drive is running for a relatively short time during the night I don't want it rotating for the rest of the time. hdparm -S 12 /dev/sdb will set the spindown to 1 minute after the last use and that works from the command line. However I would like the command to be automatically given when the drive is inserted and reading on the Internet suggests using UDEV. I've put this in a .rules file: # Set spindown time for backup drive when inserted. KERNEL=="sdb", RUN+="/sbin/hdparm -S 12 /dev/sdb" But that doesn't work i.e. nothing happens. When I first tried I put only one = after KERNEL and UDEV complained so I know it is being 'seen' and the drive is always 'sdb'. Should it work? -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain

1 0

Colour Laserjet printer problem
by Barry Samuels 25 Jan '13

25 Jan '13

I have an HP Colour LaserJet 4500 and recently it displayed a message to the effect that the black (K) toner was low. I have configuration set to continue until toner runs out but before that happened it started to show grey bands vertically down the page. I decided to change the toner in the hope that it might cure the problem but it didn't. I can see rings of toner around the imaging drum if I take it out. It does seem to be picking up surplus toner but I've no idea where from. At present I have: Transfer Kit showing 96% life left Fuser Kit showing 91% life left Drum Kit 68% lif left Any suggestions would be appreciated. If you suggest cleaning something would you please explain where to find it because I probably won't know. -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain

2 1

MySQL select into
by Mark Rogers 22 Jan '13

22 Jan '13

Any MySQL gurus here? I have a table: `item_uid` INT (Primary key) `cat_uid` INT <several other fields> What I want to do is take all records where cat_uid=10, and insert them back into the table with cat_uid=11 (in other words duplicating "category 10" as "category 11". What I had was: CREATE TEMPORARY TABLE tmp SELECT * FROM myTable WHERE cat_uid=10; UPDATE tmp SET cat_uid=11; INSERT INTO myTable SELECT * FROM tmp; This used to work when I didn't have the primary key `uid`, but now this just attempts to insert duplicates for the uid field and fails. The following works instead: CREATE TEMPORARY TABLE tmp SELECT * FROM myTable WHERE cat_uid=10; ALTER TABLE tmp CHANGE uid uid INT(11) NULL DEFAULT NULL; UPDATE tmp SET uid=NULL, cat_uid=11; INSERT INTO myTable SELECT * FROM tmp; In other words, alter the tmp table structure to allow the uid field to be NULL, and as part of my update set all the uid's to NULL, so that the final insert tries to insert NULL for uid which causes MySQL to insert the next auto-increment value instead. But this seems very convoluted, is there a "correct" (or just "better") way? -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG

2 4

Summat up with Mint - or Mozilla
by Anthony Anson 21 Jan '13

21 Jan '13

Using the most recent Thunderbird, ditto Firefox, the system freezes frequently when I have T/bird and F/fox running at the same time. F'rinstance today, I left T/bird running while I went to the BBC website. There I filled in an online comment and then tried to close Firefox, which declined to play nicely, and none of the usual rollover dialogue boxes on Mint functions worked - connection speed, state of battery etc just sat there glumly ignoring the mouse pointer. File and Quit on the function bar closed Firefox, but nothing would shift Thunderbird. Three-fingered salute deleted the last two e-mailsin my inbox... Had to power-down to regain control, but the e-mails seem to have vanished entirely. (Not in trash, nor in any other folder.) Anyone any bright ideas? (Apart from not running Firefox and Thunderbird at the same time...) -- Tony

2 4

?Ubuntu 12.10 Boot Issues
by steve-ALUG＠hst.me.uk 21 Jan '13

21 Jan '13

Hi all, I was on an old version of Ubuntu, started getting messages about it not being supported any more and so decided to update. I decided to install the lubuntu-desktop package, then do updates until I was up-to-date. So now, I'm up to Lubuntu 12.10. When I booted it, it hang big-time before getting into the GUI. I Hard-booted and selected a different kernel, and it booted fine. After experimentation it seems that booting into the kernel described as "Ubuntu 12.10, kernel 3.5.0-10-generic" causes the machine to hang before getting into the GUI. I've just tried making LUbuntu and XUbuntu 12.10 CDs and booting from them, but both of them hang, I presume when trying to load the Kernel. Trying to boot via the "kernel 3.5.0-10-generic recovery mode" boot menu option also hangs. I've not had a problem like this before. Does anyone have any advice on diagnosing what's causing this hang? Any logs to examine would have to survive a hard-reboot, or are there any debug options I can put on the grub boot line. Or am I doomed, or have to wait for a new kernel to come out, or keep using the old one. Any advice or comments would be appreciated. I haven't googled much as I don't quite know what to search for (you may tell me I'm a numpty for this!) Hardware is a Dell Lattitude D410. Cheers Steve

3 5

[ALUG Announce] 3rd Monday Monthly Ipswich Social Meeting : The Mulberry Tree
by samwise 21 Jan '13

21 Jan '13

ALUG Announce Please send announcements to Announce(a)lists.alug.org.uk Unsub, change settings via http://lists.alug.org.uk/mailman/listinfo/announce ----- The monthly "Third Monday" ALUG Ipswich Social Meeting is back tonight, 21st January, at 20:00: The Mulberry Tree 5 Woodbridge Road IPSWICH IP4 2EA Tel: 01473 225776 Map/Directions URI: http://maps.google.co.uk/?q=IP4+2EA WWW: https://www.facebook.com/MULBERRYHOSPITALITY The Mulberry Tree, previously known as The Milestone Beer House, is a grade 2 listed building in the centre of Ipswich. It is located at the intersection between Woodbridge Road and the A1156, across the road from the old Odeon cinema and not far from the Regent Theatre. It re-opened on Thursday March 29th 2012 after a major refurbishment, under its original name of The Mulberry House and has already begun to regain a good reputation for its selection of beers and ales. It has a sizeable car park to the rear of the building. All are welcome, novices and experienced users. Feel free to ask for further information on the Main ALUG mailing list. I'm back home and will be out to play today, if anyone wants to join me. :) Cheers, Peter. ----- Send discussion replies to main(a)lists.alug.org.uk (Reply-To set) Unsub, change settings via http://lists.alug.org.uk/mailman/listinfo/announce

1 0

Help with hardware
by Peter Alcibiades 20 Jan '13

20 Jan '13

Any ideas what this might be? I was having a crash with instant power down. So I replaced the PSU with a Xilence, and for a while all seemed well. Then it happened again. Restarted. But to restart, I had to actually unplug the power cable. Only then did the power switch work. All carried along fine for a week or so - I never turn it off, and ir ran just fine. Then just now, it happens again! In the meantime afflicted by a sudden rush of nerves I have made a backup using clonezilla so its not going to be the end of the world if it goes south. What happens is, try to use the power switch and nothing whatever happens. Then I connected up to a different power cable and a different socket, and it goes on. The Gigabyte board asks me to pick the last good boot, so I do. Then it powers down and restarts, and goes into the usual Debian login, after which it runs the disk diagnostic and repairs itself, and now its running just fine again. What on earth could it be and what should I do? See what is weird is that I am tempted to think maybe its the extension block, but the screen runs off that, and that is just fine. It seems reasonable to change out the extension block because they are not expensive. But I can't quite see how that could do it. There is no doubt that last time it was the psu. It was totally dead, and replacing that first with the evo and then with the xilence, it did actually bring it all back to life. But this time its clearly not the psu, or could it be? Al

3 4

A conundrum to boot!
by Barry Samuels 19 Jan '13

19 Jan '13

We have a spare computer laying around the place, as you do, which serves as a quick substitute if either of our main computers goes wrong. It's just a case of taking the disc drive out of the faulty computer and putting it in the spare and off we go again. The mainboard in the spare went wrong a while ago and I managed to get a free replacement from my brother in the form of a discarded mainboard. This mainboard is an Intel DP965LT which has a respectable 2.7 GHz Core2 Duo CPU and 8 GBs of RAM. The PSU in my main (normal) computer exploded recently quite spectacularly with a flash, bang and a puff of smoke so I thought I would swap the drive into the spare computer while I sorted things out. It was then that I discovered that although the spare board had SATA connections it wouldn't boot a SATA drive. With just the SATA drive connected I got a "No bootable drives" message. It will, however, boot an IDE drive. I do have a small IDE drive and I was wondering if it were possible to write a Lilo boot sector on that drive in order to start the Linux system on the SATA drive. But with the two drives connected the IDE drive comes up as sda and the SATA drive as sdb and, of course, the system on the SATA drive is configured to use sda. I don't want to change the configuration on the SATA drive so is there a way round this apart from trying to cadge another mainboard from somewhere? -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain

6 13

zhead/ztail
by Mark Rogers 18 Jan '13

18 Jan '13

I want to display the first few line of each of a batch of log files, some of which are gzipped. Normally I'd use zcat in place of cat, zgrep in place of grep, etc, to work with a mixture of compressed and uncompressed files. But zhead and ztail don't exist. What's the best way to achieve this: zhead -n1 /var/log/syslog* (show the first line of each syslog* file)? All the thoughts I had got increasingly convoluted. Obviously stuff like: zcat /var/log/syslog* | head -n1 .. doesn't achieve what I want... -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

2 2

Re: [ALUG] Help with hardware
by Peter Alcibiades 16 Jan '13

16 Jan '13

Yes, I did check out the loading on the various calculators I found - in response to the suggestions - and there seems to be plenty of margin. More than the calculators suggest. Its 250W, and its only a main board, one hard drive and i3. No graphics card. The earlier psu that seems to have burned out was 300W. I'm feeling doubtful that an inadequate psu can be the cause. Didn't note exactly what the restart message was. Something about going back to the last good configuration. No, I don't think there are two bios files installed. But there were a bunch of choices which I don't exactly remember. What I did so far was to take the supply off the surge protected strip and connect it directly to the mains and everything is running fine now. One thing that makes me suspect the psu is that the crash happened when watching youtube - which presumably is more processor intensive than just web or word processing. Anyway, I ordered a new extension strip, this one brand named so maybe it will be better, and we will see. I truly hate intermittent faults! Don't we all? Al On Tuesday 15 January 2013 08:01:00 Wayne Stallwood wrote: > As an educated guess I'd say either there has been a power > spike/overload or PSU failure that puts it into protection mode or there > is something going on with your mainboard. > > When you say "The Gigabyte board asks me to pick the last good boot" > what exactly is the message ? Is it falling back to a redundant BIOS image ? > > Is the PSU adequately sized for the machine..I know we had this > conversation late last year on the list. But specifically if you use a > PSU calculator for your combination of hardware components does the > current draw on the 12v rails come in at way under whatever it says for > that rail on the PSU spec. > > If you go by wattage alone you don't know if the distribution of > capacity is sufficient across the rails for your hardware. You often > can't for example draw the total rated capacity from one rail. Sometimes > despite the overall rating appearing to be sufficient the 12V rail can > be undersized, this is particularly true on PSU's designed for older > hardware. > > > > > > >

1 0

Help, graphics broken!
by Mark Rogers 16 Jan '13

16 Jan '13

Kubuntu 12.10 on a new PC, has been working fine. Just had a power cut, and on restarting I see the BIOS on my screen then it goes grey and stays that way. I can tell it's booting up (combination of HDD lights etc, and the fact that if I wait long enough I can SSH in to it). But nothing on the screen. It's possible, but no means certain, that I had does some apt-get upgrade's (or dist-upgrade's) prior to the power cut that have now come into effect, How do I debug this to get my desktop back (by which I mean visible to me - I'm sure it's there!)? In case it helps: $ lspci | grep VGA 00:01.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI BeaverCreek [Radeon HD 6530D] It's just a default install, no graphics tweaking (I haven't even installed proprietary drivers). Monitor is Edge EF220A (1920 x 1080). Mark -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG

1 1

Re: [ALUG] Leetkey
by Anthony Anson 14 Jan '13

14 Jan '13

On 14/01/13 17:39, Anthony Anson wrote: > On 13/01/13 18:07, steve-ALUG(a)hst.me.uk wrote: >> On 13/01/13 12:58, Anthony Anson wrote: >>> On 13/01/13 01:04, steve-ALUG(a)hst.me.uk wrote: >>>> On 12/01/13 12:07, Anthony Anson wrote: >>>> [SNIP] >>>>> I must say I dislike the current offering of Thunderbird. >>>>> >>>> Well I did too, until I found the various extensions and >>>> customisationshe >>>> that made it look, more or less, like it used to look like. What don't >>>> you like? >>> >>> How long have I got? I haven't (yet?) found a way to close mail on the >>> screen without clicking 'inbox'; > >> Not quite sure what you mean. > > Well, the old Thunderbird on the Eee, and Eee bah gum, it were old... > had a close box on the pane - the only 'close' box closes Thunderbird. > > Shall play with dragging functions about as you suggest - there's > already a backward and a forward arrow - ta. > >>> the only way to date that I've found to >>> go on to the next message is to remember to position the cursor in the >>> right area of the pane and use the 'down' key; > >> You can read through emails in the same style as a newsgroup reader >> program, by pressing space. Space shows you a pageful of an email at a >> time. When it gets to the end of the email, it then moves on to the >> next. >> Also, you can right click on the toolbar at the top of the screen (and >> possibly the one on top of the email), and add a Next and Previous >> button. This allows you to scroll through your emails by clicking on >> them. There's a bunch of shortcuts too - see: >> >> https://support.mozillamessaging.com/en-US/kb/keyboard-shortcuts#w_moving-a… >> >> >> >> Loads of useful stuff on that page - I've just learnt some stuff! :-) > > I shall look at it next time I can use wifi >> >>> I haven't yet found a way >>> to turn off red worms - er - spellcheck-as-you-trype, and lots more. >> >> Edit/Preferences >> Composition icon >> Spelling Tab >> Tick or untick as you see fit. >>> >>> Mind you, for reasons of my ISP's change of security and methods of >>> connecting, I haven't been able to use POP3 between getting this >>> new(ish) box before Christmas and last week, so I wasn't using the >>> program, and, I haven't tried *really* hard. >>> >>> I was considering overwriting Thunderguts on this box with the old >>> version I had been using on the Eee, but now the new one can get into >>> the mail server I'm loath to poke it with sharp sticks. >> >> as to why you're getting sent email in your inbox, well that's not a >> problem I'm getting. Is something somewhere BCCing your email back to >> you? Could it be a server/ISP issue? Does it happen when you send >> email by other means? > > I don't send email by any other means. > >> Don't get me wrong - I think Thunderbird is heading the wrong way. I've >> just installed it for the other half, and the menu bar was hidden and I >> was presented with all sorts of horrible menus and stuff, and by >> default, it looks horrible, but it is possible to wrangle it back to >> working OK with judicious use of extensions and customising. I've >> managed it! > > Thanks - working my way through all this. Worming powder applied. I had > already adjusted the menu bar, but became pi^h^h^somewhat incensed by > the unintuitive nature of the controls - it's rather like learning to > fly an airliner which is coming in to land without the help of George. > -- Tony

1 0

A cry for help
by nev young 14 Jan '13

14 Jan '13

well for hardware. Does any one have a graphics card, with a PCIE connector and SVGA D-connector on the output, going spare that Vickie could have? If so could they bring it to the ALUG meet tomorrow. I've taken one out of my (spit)windows PC to get hers going and would like to avoid having to part with cash for a new one. Thankx Nev

2 3

Leetkey
by Anthony Anson 13 Jan '13

13 Jan '13

ISTR a workround for adding Leetkey to Thunderbird 17 was sent to the list. Am I dreaming? If my memory is correct I can rootle through everything I have saved from ALUG on the old Eee. If I am wrong, what is the last version in which Leetkey works? If I go onto the Mozilla website and install it, will it replace Mozilla 17, or install another one beside it? -- Tony

4 8

Backup Disk Freeze - Any cause clue from syslog?
by steve-alug＠hst.me.uk 13 Jan '13

13 Jan '13

Hi! My backup system has recently started causing me some grief. I'm running backuppc backing up to an IDE drive in a Netgear SC101 SAN storage box. There are two drives in this device, and the second one is being talked to by Simple Backup, and is working fine. The main/bigger drive is /dev/nbd1 and backuppc has started hanging whilst using it. It doesn't happen straight away - it's several hours after the backup has been started. If I power everything down, wait and switch it back on again, all the drives spin up and seem to work ok again. The SC101 is attached via the SC101-nbd project driver. http://code.google.com/p/sc101-nbd/ It has all been working fine until recently. Nothing's changed, apart from installing routine security updates (inc kernel updates) I'm running ubuntu 10.04. I'm currently doing a deep check on the disk using a tool from Ultimate boot Cd appropriate for the drive manufacturer, but a quick check said the disk's OK. Any ideas why it's hanging like this? Once hung, I can't even ls the disk - it's completely blocked. I tried reformatting it yesterday, and erasing and reinstalling backuppc, but it crashed again. Some of Syslog output below - there's many more after this one, but this is the first. Possible candidates: Disk SAN controller Main PC faulty Disk sleeping and failing to wake again. Any ideas guys? Cheers Steve Jan 11 07:00:28 myserver ntpd[1950]: kernel time sync status change 6001 Jan 11 07:02:01 myserver CRON[11758]: (logcheck) CMD ( if [ -x /usr/sbin/logcheck ]; then nice -n10 /usr/sbin/logcheck; fi) Jan 11 07:04:11 myserver kernel: [45960.212092] INFO: task jbd2/nbd1-8:12704 blocked for more than 120 seconds. Jan 11 07:04:11 myserver kernel: [45960.212100] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Jan 11 07:04:11 myserver kernel: [45960.212104] jbd2/nbd1-8 D 00005645 0 12704 2 0x00000000 Jan 11 07:04:11 myserver kernel: [45960.212113] c6ce9de4 00000046 f7490000 00005645 00000000 c08964e0 c47b9c24 c08964e0 Jan 11 07:04:11 myserver kernel: [45960.212123] 3c1ac4f3 0000299c c08964e0 c08964e0 c47b9c24 c08964e0 c08964e0 d7e7d800 Jan 11 07:04:11 myserver kernel: [45960.212131] 00000000 0000299c c47b9980 c47b9980 c30094e0 c6ce9e2c c6ce9df4 c05b71a1 Jan 11 07:04:11 myserver kernel: [45960.212140] Call Trace: Jan 11 07:04:11 myserver kernel: [45960.212157] [<c05b71a1>] io_schedule+0x61/0xa0 Jan 11 07:04:11 myserver kernel: [45960.212165] [<c01d4b9d>] sync_page+0x3d/0x50 Jan 11 07:04:11 myserver kernel: [45960.212171] [<c05b795d>] __wait_on_bit+0x4d/0x70 Jan 11 07:04:11 myserver kernel: [45960.212176] [<c01d4b60>] ? sync_page+0x0/0x50 Jan 11 07:04:11 myserver kernel: [45960.212181] [<c01d4dc1>] wait_on_page_bit+0x91/0xa0 Jan 11 07:04:11 myserver kernel: [45960.212187] [<c0170e90>] ? wake_bit_function+0x0/0x50 Jan 11 07:04:11 myserver kernel: [45960.212193] [<c01d5051>] wait_on_page_writeback_range+0xa1/0x110 Jan 11 07:04:11 myserver kernel: [45960.212201] [<c023ca2e>] ? bio_alloc_bioset+0x2e/0xf0 Jan 11 07:04:11 myserver kernel: [45960.212209] [<c02da977>] ? jbd2_journal_write_metadata_buffer+0x277/0x370 Jan 11 07:04:11 myserver kernel: [45960.212215] [<c01d511a>] filemap_fdatawait+0x5a/0x70 Jan 11 07:04:11 myserver kernel: [45960.212221] [<c02d2bae>] jbd2_journal_commit_transaction+0x51e/0x1030 Jan 11 07:04:11 myserver kernel: [45960.212230] [<c01311b8>] ? default_spin_lock_flags+0x8/0x10 Jan 11 07:04:11 myserver kernel: [45960.212236] [<c0164c78>] ? try_to_del_timer_sync+0x68/0xb0 Jan 11 07:04:11 myserver kernel: [45960.212241] [<c02d93d5>] kjournald2+0x95/0x1c0 Jan 11 07:04:11 myserver kernel: [45960.212247] [<c0170e40>] ? autoremove_wake_function+0x0/0x50 Jan 11 07:04:11 myserver kernel: [45960.212252] [<c02d9340>] ? kjournald2+0x0/0x1c0 Jan 11 07:04:11 myserver kernel: [45960.212256] [<c0170bb4>] kthread+0x74/0x80 Jan 11 07:04:11 myserver kernel: [45960.212261] [<c0170b40>] ? kthread+0x0/0x80 Jan 11 07:04:11 myserver kernel: [45960.212267] [<c010a467>] kernel_thread_helper+0x7/0x10 Jan 11 07:04:11 myserver kernel: [45960.212278] INFO: task flush-43:16:11670 blocked for more than 120 seconds. Jan 11 07:04:11 myserver kernel: [45960.212281] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Jan 11 07:04:11 myserver kernel: [45960.212285] flush-43:16 D 00000484 0 11670 2 0x00000000 Jan 11 07:04:11 myserver kernel: [45960.212291] c46d3b34 00000046 c07eaa20 00000484 00000000 c08964e0 ef5735a4 c08964e0 Jan 11 07:04:11 myserver kernel: [45960.212300] f5f3884b 0000299a c08964e0 c08964e0 ef5735a4 c08964e0 c08964e0 e8566a00 Jan 11 07:04:11 myserver kernel: [45960.212308] f5f2b230 0000299a ef573300 c2e0afe0 eb6accc0 c46d3b74 c46d3b94 c02d1ec5 Jan 11 07:04:11 myserver kernel: [45960.212316] Call Trace: Jan 11 07:04:11 myserver kernel: [45960.212324] [<c02d1ec5>] do_get_write_access+0x245/0x440 Jan 11 07:04:11 myserver kernel: [45960.212330] [<c0170e90>] ? wake_bit_function+0x0/0x50 Jan 11 07:04:11 myserver kernel: [45960.212336] [<c02d2238>] jbd2_journal_get_write_access+0x28/0x40 Jan 11 07:04:11 myserver kernel: [45960.212343] [<c02bb3fe>] __ext4_journal_get_write_access+0x2e/0x60 Jan 11 07:04:11 myserver kernel: [45960.212349] [<c02bd611>] ext4_mb_mark_diskspace_used+0x61/0x2f0 Jan 11 07:04:11 myserver kernel: [45960.212354] [<c02c38a6>] ext4_mb_new_blocks+0x306/0x590 Jan 11 07:04:11 myserver kernel: [45960.212360] [<c02b8613>] ? ext4_ext_find_extent+0x253/0x280 Jan 11 07:04:11 myserver kernel: [45960.212365] [<c02baf12>] ext4_ext_get_blocks+0x472/0x700 Jan 11 07:04:11 myserver kernel: [45960.212373] [<c0298f98>] ext4_get_blocks+0xf8/0x2c0 Jan 11 07:04:11 myserver kernel: [45960.212379] [<c02997c9>] mpage_da_map_blocks+0x99/0x410 Jan 11 07:04:11 myserver kernel: [45960.212385] [<c02b0d90>] ? ext4_journal_start_sb+0xc0/0xf0 Jan 11 07:04:11 myserver kernel: [45960.212390] [<c0296917>] ? ext4_meta_trans_blocks+0xb7/0xd0 Jan 11 07:04:11 myserver kernel: [45960.212396] [<c0299dc9>] ext4_da_writepages+0x289/0x580 Jan 11 07:04:11 myserver kernel: [45960.212403] [<c01489d9>] ? update_curr+0x169/0x2a0 Jan 11 07:04:11 myserver kernel: [45960.212410] [<c01dc66c>] do_writepages+0x1c/0x40 Jan 11 07:04:11 myserver kernel: [45960.212417] [<c0232a15>] writeback_single_inode+0xd5/0x380 Jan 11 07:04:11 myserver kernel: [45960.212422] [<c02331f9>] writeback_sb_inodes+0x149/0x230 Jan 11 07:04:11 myserver kernel: [45960.212427] [<c023364a>] writeback_inodes_wb+0x8a/0x160 Jan 11 07:04:11 myserver kernel: [45960.212432] [<c0233903>] wb_writeback+0x1e3/0x230 Jan 11 07:04:11 myserver kernel: [45960.212438] [<c0233a81>] wb_do_writeback+0x131/0x140 Jan 11 07:04:11 myserver kernel: [45960.212443] [<c0233acb>] bdi_writeback_task+0x3b/0xb0 Jan 11 07:04:11 myserver kernel: [45960.212448] [<c01eb9e0>] ? bdi_start_fn+0x0/0xc0 Jan 11 07:04:11 myserver kernel: [45960.212453] [<c01eba3d>] bdi_start_fn+0x5d/0xc0 Jan 11 07:04:11 myserver kernel: [45960.212457] [<c01eb9e0>] ? bdi_start_fn+0x0/0xc0 Jan 11 07:04:11 myserver kernel: [45960.212461] [<c0170bb4>] kthread+0x74/0x80 Jan 11 07:04:11 myserver kernel: [45960.212466] [<c0170b40>] ? kthread+0x0/0x80 Jan 11 07:04:11 myserver kernel: [45960.212471] [<c010a467>] kernel_thread_helper+0x7/0x10 Jan 11 07:05:01 myserver CRON[11897]: (root) CMD (if [ -x /usr/bin/vnstat ] && [ `ls /var/lib/vnstat/ | wc -l` -ge 1 ]; then /usr/bin/vnstat -u; fi) Jan 11 07:06:11 myserver kernel: [46080.212093] INFO: task jbd2/nbd1-8:12704 blocked for more than 120 seconds. Jan 11 07:06:11 myserver kernel: [46080.212101] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Jan 11 07:06:11 myserver kernel: [46080.212106] jbd2/nbd1-8 D 00005645 0 12704 2 0x00000000 Jan 11 07:06:11 myserver kernel: [46080.212114] c6ce9de4 00000046 f7490000 00005645 00000000 c08964e0 c47b9c24 c08964e0 Jan 11 07:06:11 myserver kernel: [46080.212124] 3c1ac4f3 0000299c c08964e0 c08964e0 c47b9c24 c08964e0 c08964e0 d7e7d800 Jan 11 07:06:11 myserver kernel: [46080.212132] 00000000 0000299c c47b9980 c47b9980 c30094e0 c6ce9e2c c6ce9df4 c05b71a1 Jan 11 07:06:11 myserver kernel: [46080.212140] Call Trace: Jan 11 07:06:11 myserver kernel: [46080.212156] [<c05b71a1>] io_schedule+0x61/0xa0 Jan 11 07:06:11 myserver kernel: [46080.212165] [<c01d4b9d>] sync_page+0x3d/0x50 Jan 11 07:06:11 myserver kernel: [46080.212171] [<c05b795d>] __wait_on_bit+0x4d/0x70 Jan 11 07:06:11 myserver kernel: [46080.212175] [<c01d4b60>] ? sync_page+0x0/0x50 Jan 11 07:06:11 myserver kernel: [46080.212180] [<c01d4dc1>] wait_on_page_bit+0x91/0xa0 Jan 11 07:06:11 myserver kernel: [46080.212187] [<c0170e90>] ? wake_bit_function+0x0/0x50 Jan 11 07:06:11 myserver kernel: [46080.212192] [<c01d5051>] wait_on_page_writeback_range+0xa1/0x110 Jan 11 07:06:11 myserver kernel: [46080.212200] [<c023ca2e>] ? bio_alloc_bioset+0x2e/0xf0 Jan 11 07:06:11 myserver kernel: [46080.212207] [<c02da977>] ? jbd2_journal_write_metadata_buffer+0x277/0x370 Jan 11 07:06:11 myserver kernel: [46080.212213] [<c01d511a>] filemap_fdatawait+0x5a/0x70 Jan 11 07:06:11 myserver kernel: [46080.212219] [<c02d2bae>] jbd2_journal_commit_transaction+0x51e/0x1030 Jan 11 07:06:11 myserver kernel: [46080.212228] [<c01311b8>] ? default_spin_lock_flags+0x8/0x10 Jan 11 07:06:11 myserver kernel: [46080.212234] [<c0164c78>] ? try_to_del_timer_sync+0x68/0xb0 Jan 11 07:06:11 myserver kernel: [46080.212239] [<c02d93d5>] kjournald2+0x95/0x1c0 Jan 11 07:06:11 myserver kernel: [46080.212245] [<c0170e40>] ? autoremove_wake_function+0x0/0x50 Jan 11 07:06:11 myserver kernel: [46080.212250] [<c02d9340>] ? kjournald2+0x0/0x1c0 Jan 11 07:06:11 myserver kernel: [46080.212254] [<c0170bb4>] kthread+0x74/0x80 Jan 11 07:06:11 myserver kernel: [46080.212259] [<c0170b40>] ? kthread+0x0/0x80 Jan 11 07:06:11 myserver kernel: [46080.212265] [<c010a467>] kernel_thread_helper+0x7/0x10 Jan 11 07:06:11 myserver kernel: [46080.212277] INFO: task flush-43:16:11670 blocked for more than 120 seconds. Jan 11 07:06:11 myserver kernel: [46080.212282] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Jan 11 07:06:11 myserver kernel: [46080.212285] flush-43:16 D 00000484 0 11670 2 0x00000000 Jan 11 07:06:11 myserver kernel: [46080.212292] c46d3b34 00000046 c07eaa20 00000484 00000000 c08964e0 ef5735a4 c08964e0 Jan 11 07:06:11 myserver kernel: [46080.212300] f5f3884b 0000299a c08964e0 c08964e0 ef5735a4 c08964e0 c08964e0 e8566a00 Jan 11 07:06:11 myserver kernel: [46080.212309] f5f2b230 0000299a ef573300 c2e0afe0 eb6accc0 c46d3b74 c46d3b94 c02d1ec5 Jan 11 07:06:11 myserver kernel: [46080.212317] Call Trace: Jan 11 07:06:11 myserver kernel: [46080.212325] [<c02d1ec5>] do_get_write_access+0x245/0x440 Jan 11 07:06:11 myserver kernel: [46080.212331] [<c0170e90>] ? wake_bit_function+0x0/0x50 Jan 11 07:06:11 myserver kernel: [46080.212337] [<c02d2238>] jbd2_journal_get_write_access+0x28/0x40 Jan 11 07:06:11 myserver kernel: [46080.212344] [<c02bb3fe>] __ext4_journal_get_write_access+0x2e/0x60 Jan 11 07:06:11 myserver kernel: [46080.212349] [<c02bd611>] ext4_mb_mark_diskspace_used+0x61/0x2f0 Jan 11 07:06:11 myserver kernel: [46080.212354] [<c02c38a6>] ext4_mb_new_blocks+0x306/0x590 Jan 11 07:06:11 myserver kernel: [46080.212359] [<c02b8613>] ? ext4_ext_find_extent+0x253/0x280 Jan 11 07:06:11 myserver kernel: [46080.212364] [<c02baf12>] ext4_ext_get_blocks+0x472/0x700 Jan 11 07:06:11 myserver kernel: [46080.212372] [<c0298f98>] ext4_get_blocks+0xf8/0x2c0 Jan 11 07:06:11 myserver kernel: [46080.212377] [<c02997c9>] mpage_da_map_blocks+0x99/0x410 Jan 11 07:06:11 myserver kernel: [46080.212383] [<c02b0d90>] ? ext4_journal_start_sb+0xc0/0xf0 Jan 11 07:06:11 myserver kernel: [46080.212388] [<c0296917>] ? ext4_meta_trans_blocks+0xb7/0xd0 Jan 11 07:06:11 myserver kernel: [46080.212393] [<c0299dc9>] ext4_da_writepages+0x289/0x580 Jan 11 07:06:11 myserver kernel: [46080.212400] [<c01489d9>] ? update_curr+0x169/0x2a0 Jan 11 07:06:11 myserver kernel: [46080.212407] [<c01dc66c>] do_writepages+0x1c/0x40 Jan 11 07:06:11 myserver kernel: [46080.212413] [<c0232a15>] writeback_single_inode+0xd5/0x380 Jan 11 07:06:11 myserver kernel: [46080.212418] [<c02331f9>] writeback_sb_inodes+0x149/0x230 Jan 11 07:06:11 myserver kernel: [46080.212422] [<c023364a>] writeback_inodes_wb+0x8a/0x160 Jan 11 07:06:11 myserver kernel: [46080.212427] [<c0233903>] wb_writeback+0x1e3/0x230 Jan 11 07:06:11 myserver kernel: [46080.212433] [<c0233a81>] wb_do_writeback+0x131/0x140 Jan 11 07:06:11 myserver kernel: [46080.212438] [<c0233acb>] bdi_writeback_task+0x3b/0xb0 Jan 11 07:06:11 myserver kernel: [46080.212443] [<c01eb9e0>] ? bdi_start_fn+0x0/0xc0 Jan 11 07:06:11 myserver kernel: [46080.212447] [<c01eba3d>] bdi_start_fn+0x5d/0xc0 Jan 11 07:06:11 myserver kernel: [46080.212451] [<c01eb9e0>] ? bdi_start_fn+0x0/0xc0 Jan 11 07:06:11 myserver kernel: [46080.212455] [<c0170bb4>] kthread+0x74/0x80 Jan 11 07:06:11 myserver kernel: [46080.212459] [<c0170b40>] ? kthread+0x0/0x80 Jan 11 07:06:11 myserver kernel: [46080.212464] [<c010a467>] kernel_thread_helper+0x7/0x10 Jan 11 07:08:11 myserver kernel: [46200.212124] INFO: task jbd2/nbd1-8:12704 blocked for more than 120 seconds.

3 3

Have I bought the wrong kit?
by Chris Walker 11 Jan '13

11 Jan '13

I am fed up with the cheap Huawei router that Talktalk supply so decided I would go for something that will run ddwrt. Having looked at the ddwrt site, I then started looking at Amazon. I decided I would buy a Buffalo Air Station and went for a WZR-HP-AG300H. But it tells me I should connect it to my modem. <Thinks> Hmm. Thought that was built-in. Look again at spec. B*gger. So should I buy a cheap ADSL modem or ask Amazon if they'll take it back and buy another Air Station but this time, one with the correct specs?

2 1

Barebones
by Laurie Brown 11 Jan '13

11 Jan '13

Over the last few years, we have been building small servers using Asus barebones systems. They've been great, and we can install whatever CPU we like, disks, memory, flavour of Linux and so on. They have proven to be relatively cheap and highly reliable machines. However, I just can't seem to find them any more! Our requirement is room for at least 3 SATA disks, preferably 4 plus DVD, multi-core CPU, preferably AMD, good PSU, and capable of taking at least 8Gb of memory. I can find loads of tiny machines great for firewalls and so on, but nothing any larger. Even the Shuttles, still easily found, can't hold enough disks. Does anyone know of a reliable supplier of Asus barebones, or something similar? The last thing I want to do is spend 4 figures on an HP and throw away a licence for Winblow$. Yes, I know I can buy a case, mobo and so on, but that's a laborious exercise if regular, and kit becomes obsolete so fast, it has to be repeated almost every time we build a machine. We've looked at Zoostorm, which is the closest I can find, but they aren't barebones, and we don't need a 2tb disk, just because we want 8Gb of memory. With Zoostorm, it seems that the higher the CPU/memory spec, the bigger the disks, and so on, and it's very hard to find out what the mobo is, and whether it's Linux-friendly. Just too many unknowns and potentially expensive pitfalls. Help/guidance greatly appreciated! Cheers, Laurie. -- --------------------------------------------------------------------- Laurie Brown laurie(a)brownowl.com ---------------------------------------------------------------------

4 7

Going spare...
by Mark Rogers 11 Jan '13

11 Jan '13

I have this motherboard, with CPU (couldn't tell you what CPU from memory): http://reviews.cnet.com/motherboards/abit-nf7-s2g-motherboard/4505-3049_7-3… Could likely find some RAM to go with it. Socket A so not high spec, but might be a good replacement for a failed board. I also have at least one Fujitsu Siemens Econel server with dead motherboard (PSU and CPU fine, I believe). Good solid metal cases with loads of drive bays etc, but not worth repairing. Should take standard motherboards if you can find one that the CPU fits relative to where the PSU is. Free for collection from Peterborough. Mark -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG

1 0

Nice confirmation of mail not getting lost when SMTP server isn't available
by Chris Green 11 Jan '13

11 Jan '13

Over the past few days I have been playing about with the DNS on my LAN (changing which server runs dnsmasq, etc.). Last night around 22:00 I left everything, as I thought, running OK. This morning though no E-Mails had arrived on my desktop (which is also the SMTP server). It turned out that I hadn't restarted things enough and my dekstop had got a randomly assigned IP address rather than the fixed one it should have got (by telling dnsmasq its MAC address). So TsoHost's mail system hadn't been able to deliver mail to zbmc.eu for something like 12 hours. It was a good test for two things:- My offsite mail backup to which I deliver all my mail in parallel to delivery to my dekstop worked perfectly, so I could see all the undelivered mail there. When I fixed the IP address of my desktop all the undelivered mail arrived on my desktop after an hour or two. Isn't it nice when things work as they should! :-) -- Chris Green

1 0

Anyone use x2go or similar software?
by Chris Green 11 Jan '13

11 Jan '13

I was considering using VNC to access another box on my LAN but as the remote system (which would run a VNC server) doesn't currently have X running on it (though there's a Lubuntu desktop available) it makes VNC rather awkward. I've seen x2go recommended for this sort of situation as it seems as if it's more oriented towards the sort of thing I want to do - fire up an X session and use it remotely. While it would be nice if one could use 'old fashioned' X as it was originally designed it tends not to work so well with very many programs now using dBus to communicate. E.g. if I try and run nm-connection-editor on the remote system via an "ssh -x remote-system" connection it pops up a window on my system but there's an error:- ** (nm-connection-editor:3088): WARNING **: Couldn't connect to accessibility bus: Failed to connect to socket /tmp/dbus-9FFbE3UfRz: Connection refused ... and it doesn't actually work. Or can anyone suggest how to get around the dBus communication issue? -- Chris Green

3 4

Full size but low power PCs - does nobody sell them? (maybe related to Laurie's recent question)
by Chris Green 10 Jan '13

10 Jan '13

Why is it that no one seems to sell 'ordinary' PCs, in ordinary sized boxes, but with reasonably power efficient components. It's next to impossible to even find out what the power consumption is of most desktop PCs and very few of them have efficient power supplies (often the worst offender). It's all very well advocating little Intel Atom based boxes but they almost never have any space for extra disk drives. The same applies to Raspberry Pi etc., very low power but no easy way to assemble in a box with a couple of disk drives. I guess I'm after much the same as Laurie but with maybe more emphasis on power consumption. A basic case with low power mother board, low end processor, etc. and space for several disks. -- Chris Green

2 3

Are there any linux based NAS systems out there with reasonable audio output?
by Chris Green 10 Jan '13

10 Jan '13

I want a compact, low power Linux box to use as a NAS and audio streamer (probably running mpd). This is a slightly odd mix of requirements, does anyone have any suggestions of what to look at? Alternatively maybe a 'media streamer' would provide what I want, but I need it to have audio outputs and be able to run something like mpd so that it can run headless and be driven remotely. A custom built Linux box is a valid possibility too but getting a small, low-power one isn't trivial. (Raspberry Pi is out because its audio is notoriously bad and there seem to be few workarounds to avoid its sound problems) -- Chris Green

3 7

Re: [ALUG] Full size but low power PCs - does nobody sell them? (maybe related to Laurie's recent question)
by Chris Green 10 Jan '13

10 Jan '13

On Thu, Jan 10, 2013 at 04:47:47PM +0000, Anthony Anson wrote: > On 10/01/13 16:19, Chris Green wrote: > >Why is it that no one seems to sell 'ordinary' PCs, in ordinary sized > >boxes, but with reasonably power efficient components. > > If you don't feel competent to build one from scratch, go to a > proper computer shop (where are you? I can recommend some in the > Norwich area.) and they will put one together for you - probably > cheaper than you could. > I've been building my PCs since quite a long time ago (like since they were 386 processors, and NextGen processors - remember them?). However nowadays it doesn't really make economic sense. > >It's next to impossible to even find out what the power consumption is > >of most desktop PCs and very few of them have efficient power supplies > >(often the worst offender). > > > >It's all very well advocating little Intel Atom based boxes but they > >almost never have any space for extra disk drives. The same applies to > >Raspberry Pi etc., very low power but no easy way to assemble in a box > >with a couple of disk drives. > > Not a tremendous challenge though. > It's not a challenge putting a Raspberry Pi in a box and adding some USB disks, no. However it won't be a particularly good server. The difficulty is finding out which motherboards are low power and then finding a suitably efficient power supply. My desktop machine (Intel I3, 8Gb memory, two 1Tb disks) consumes only about 40 watts because I chose a reasonably low power board (after much searching to find information) and an efficient power supply. It's still a *bit* of a bodge though because the power supply isn't designed to fit in a standard box. Since my system only uses 40 watts it needs a much smaller power supply than most full-sized PC ones. Even a very efficient 300 watt power supply isn't going to be particularly good at 40 watts. I'm using an 80 watt supply. > >I guess I'm after much the same as Laurie but with maybe more emphasis > >on power consumption. A basic case with low power mother board, low end > >processor, etc. and space for several disks. > > Then you can run Tom's RTBT and have a low-end OS too? > > This is where your competent small trader scores. Try Anglian Internet. > I'd be *really* surprised if they know power consumption figures for assembled PCs (before they assemble them that is, I can measure the consumption of something I've built). I will send them an enquiry though. -- Chris Green

1 0

Reassembling RAID array
by Mark Rogers 10 Jan '13

10 Jan '13

I had a RAID5 array set up as /dev/md1 using 4 external USB3 drives. Following a reboot (I assume, although to be honest I'm not sure what caused it) the array hasn't come back up correctly: $ cat /proc/mdstat Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10] md1 : active raid5 super 1.2 level 5, 512k chunk, algorithm 2 [4/0] [____] md0 : active raid5 sda1[0] sdb1[1] sdd1[3] sdc1[2] 5860535808 blocks level 5, 64k chunk, algorithm 2 [4/4] [UUUU] unused devices: <none> Note: /dev/md0 is a separate array and is working fine; md1 should comprise sd[fghi]1 All the "missing" partitions are present, eg: $sudo fdisk -l /dev/sdf Disk /dev/sdf: 2000.4 GB, 2000398934016 bytes 81 heads, 63 sectors/track, 765633 cylinders, total 3907029168 sectors Units = sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x2c2d5341 Device Boot Start End Blocks Id System /dev/sdf1 2048 3907029167 1953513560 fd Linux RAID autodetect (same applies for the other three disks). How do I get mdadm to add my disks into the array, both now and after a reboot? -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

2 11

Best way to code a linux comms daemon
by Mark Rogers 09 Jan '13

09 Jan '13

Occasionally I have the need to write some communications software, typically something that connects to a serial or TCP port, writes requests and reads responses, writing the results into a database. Usually the basics are simple enough to achieve, but the challenge is in making a robust solution that will cope with lost connections and any other variety of reasons why things might go wrong (including bugs in the script itself). What is the "best" way to approach something like this? Are some programming languages better than others for this? What's the best way to ensure a script restarts if it fails for some reason? I'm looking for something Debian/Ubuntu based if relevant. [Aside: One reason for asking is that I have been playing with an app I downloaded that talks to CurrentCost power meters, with a PHP-based web interface and a Python-based daemon, and I'm not sure why Python was chosen for it; also I note that the daemon regularly crashes on my system leaving me without any data.] -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

3 6