- luv-main - example.com

GPT rescue
by Russell Coker 01 Dec '11

01 Dec '11

A client has an eSATA dock for SATA disks and he's been using a 2TB disk on his Mac. He says that he just came in one morning to see that it was claiming to be empty. When I ran the print command from parted I got the following: (parted) print Warning: /dev/sdd contains GPT signatures, indicating that it has a GPT table. However, it does not have a valid fake msdos partition table, as it should. Perhaps it was corrupted -- possibly by a program that doesn't understand GPT partition tables. Or perhaps you deleted the GPT table, and are now using an msdos partition table. Is this a GPT partition table? Yes/No? y Model: ATA eSATA-2 (scsi) Disk /dev/sdd: 2000GB Sector size (logical/physical): 512B/512B Partition Table: gpt Number Start End Size File system Name Flags 1 20.5kB 210MB 210MB fat32 EFI System Partition boot 2 210MB 2000GB 2000GB hfs+ Untitled A web search for what to do suggested creating a new label and using the rescue command: (parted) mklabel New disk label type? gpt Warning: The existing disk label on /dev/sdd will be destroyed and all data on this disk will be lost. Do you want to continue? Yes/No? y Warning: /dev/sdd contains GPT signatures, indicating that it has a GPT table. However, it does not have a valid fake msdos partition table, as it should. Perhaps it was corrupted -- possibly by a program that doesn't understand GPT partition tables. Or perhaps you deleted the GPT table, and are now using an msdos partition table. Is this a GPT partition table? Yes/No? y (parted) print Model: ATA eSATA-2 (scsi) Disk /dev/sdd: 2000GB Sector size (logical/physical): 512B/512B Partition Table: gpt Number Start End Size File system Name Flags (parted) rescue Start? 1 End? 2000G Information: A fat32 primary partition was found at 20.5kB -> 210MB. Do you want to add it to the partition table? Yes/No/Cancel? yes (parted) print Model: ATA eSATA-2 (scsi) Disk /dev/sdd: 2000GB Sector size (logical/physical): 512B/512B Partition Table: gpt Number Start End Size File system Name Flags 1 20.5kB 210MB 210MB fat32 All good so far! But after that I couldn't rescue any more. So I've got the partition I don't care about (the one the client didn't even know existed) but not the one with important data. I tried creating a new partition starting at 210MB but parted gave me one starting at 211MB. :( Does anyone know how to solve this? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

4 4

Re: mysql query
by Russell Coker 30 Nov '11

30 Nov '11

delete from horde_sessionhandler where session_lastmodified < now() - 10800; It turned out that the biggest problem I had was old session data not being cleaned up. It seems that the PHP GC doesn't seem to work as it should. Running the above command freed 3.6G of space. On Mon, 22 Aug 2011, Arjen Lentz <arjen(a)lentz.com.au> wrote: > You can do cross-database joins (dbs in mysql are just a namespace thing), > and also multi-table deletes which is an SQL extension (although you can > then only delete from one table specified in the join). delete from horde.horde_prefs,isp.users where horde.horde_prefs.pref_uid = isp.users.uid and bluebottle.users.inactive=1 and isp.users.inactive_time < now() -365*24*3600 limit 10; Would the above delete the matching rows from horde.horde_prefs without deleting from isp.users? A select with the same where clauses gets the data I want. > The horde schema is monstrous though, often the references from other > tables comprise of 3-part varchar fields - so be sure you grab the right > (and all) the relevant fields for your joins. Check with a select before > you do delete. The Horde database has 5 tables. horde_categories, horde_vfs_seq, and horde_vfs have 0, 1, and 6 rows respectively. The other two are the horde_sessionhandler table (which is only needed while I have active-active servers for the web space, I plan to go to active-passive and remove this) and the horde_prefs table. What are the inter-table references that you refer to? Thanks for your suggestions. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

1 0

IPv6 routing
by Russell Coker 26 Nov '11

26 Nov '11

I have just got the following message from my hosting provider. I have put in XXXX and YYYY to replace some of the numbers. ip -6 addr add 2a01:XXXX:YYYY:9081::2/64 dev xenbr0 I ran the above command to assign the IP address. # ping6 2a01:XXXX:YYYY:9080::1 connect: Network is unreachable But ping6 fails in the above manner. What am I doing wrong? We have successfully setup the following IPv6 subnet for your server. IPs: 2a01:XXXX:YYYY:9081:: /64 Gateway: 2a01:XXXX:YYYY:9080::1 /59 Useable IP addresses: 2a01:XXXX:YYYY:9081::2 to 2a01:XXXX:YYYY:9081:ffff:ffff:ffff:ffff -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

8 12

[luv-main] Two questions: bash script review and awk/sed/grep
by Marcus Furlong 24 Nov '11

24 Nov '11

Hi, Following on from the recent bash script reviews, if anyone would like to comment on this one, please do: https://code.vpac.org/gitorious/patchman/patchman/blobs/master/client/patch… The above script attempts to obtain a list of installed packages and enabled repos on debian/centos hosts, and sends that list to a central server. However, I'm having some sed/grep/awk issues. I'm trying to parse the output of "yum repolist enabled --verbose" and return each repo on a line as so 'rpm' 'this-repo-name' 'http://url1' 'http://url2' The output is similar to the following: Repo-id : extras Repo-name : CentOS-5 - Extras Repo-updated : Tue Oct 25 03:18:58 2011 Repo-pkgs : 237 Repo-size : 124 M Repo-baseurl : http://ftp.swin.edu.au/centos/5/extras/x86_64/, : http://ftp.monash.edu.au/pub/linux/CentOS/5/extras/x86_64/, : http://mirror.aarnet.edu.au/pub/centos/5/extras/x86_64/ Repo-expire : 3,600 second(s) (last: Wed Nov 16 12:39:48 2011) So I just need the Repo-name and the Repo-baseurls for each repo. If there is one baseurl, my version works, but if there are multiple, it breaks: # yum repolist enabled --verbose | egrep "Repo-name|Repo-baseurl" | sed -e ':a;N;$!ba;s/\nRepo-baseurl//g' -e "s/Repo-name[ ]*: /'/g" -e "s/[ ]*:[ ]\+/' '/g" | sed -e "s/$/'/g" -e "s/'/ ${host_arch}'/2" -e "s/\/'/'/g" -e "s/ ' '/' '/" 'CentOS-5 - Base' 'http://ftp.swin.edu.au/centos/5/os/x86_64/,' 'CentOS-5 - Extras' 'http://ftp.swin.edu.au/centos/5/extras/x86_64/,' 'CentOS-5 - Updates' 'http://ftp.swin.edu.au/centos/5/updates/x86_64/,' The extra comma at the end of the url is not meant to be there, and the additional baseurls are missing. Are there any awk or sed gurus that can help me get the final part of this? Thanks, Marcus. -- Marcus Furlong

3 8

pacemaker query
by James Harper 24 Nov '11

24 Nov '11

Is anyone here using pacemaker clustering? I have a question about monitoring... if I create my resource (a xen vm) in an unmanaged state initially, the monitor is never called. This is contrary to what I thought should happen, and from what I've read. If I do 'crm status ops' there is no entry for it, presumably because it has never run on a node (even though the VM is actually running). Am I doing something wrong, or is this the expected behaviour? Thanks James

1 0

pinging a down server
by Russell Coker 23 Nov '11

23 Nov '11

It's quite common for me to ping a server while it's down. Often when doing so I have a rough idea of how long it will take (for example rebooting the new Hetzner servers I'm running in Germany takes about 160 seconds during which there are no ping responses) so it's good to know how long it is before I should expect a response. At other times I want a running count of how long it's been down so I can give an accurate complaint when someone answers my call. The standard ping programs tell you the packet count (which by default is 1 per second) when it receives a response. When there is no response nothing is printed. I would like a ping type program which will tell me how many seconds the target has not been responding for. Does anyone know of such a program? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

4 4

[luv-main] this list
by Russell Coker 23 Nov '11

23 Nov '11

What's happening with the list? We need to have either luv-main(a)lists.luv.asn.au or luv-main(a)luv.asn.au become invalid to avoid duplicate messages to the list. We have had a discussion about the subject modification and so far no-one has expressed any great desire to have the subject mangled. Can we just get this fixed? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

4 6

[luv-main] Can you have too much swap?
by Leigh Sharpe 23 Nov '11

23 Nov '11

Hi All, I'm about to set up a new linux box, but not all of the bits have arrived yet. I currently only have 512MB of RAM in the machine, but ultimately, it will have at least 2GB, possibly 4. If I let the installer do it's own thing with HDD partitioning, it will set up a swap size of 2 x RAM. But that's not what I want. 2 x the current amount of RAM will be inadequate once I install more. So, if I set up 8GB of swap, expecting that I'll eventually have 4GB of RAM, is having too much swap going to be a problem in the meantime? Is there a downside to having a large amount of swap? Leigh.

10 15

[luv-main] DDNS Server [Scanned]
by Christopher M. Bailey 21 Nov '11

21 Nov '11

Hi All, I'm looking for a DDNS package that I can setup on a CentOS 5 server for potentially thousands of devices deployed at thousands of sites world wide. I have had a quick look at Bind9, which I think may buckle under this load if deployed by itself. I have also looked gnudip, but would like some feedback on how easy it is to setup and maintain, also any other packages people have used. TIA, Chris -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.

4 3

Re: [luv-main] DDNS Server [Scanned]
by Christopher M. Bailey 21 Nov '11

21 Nov '11

----- Original Message ----- From: "Trent W. Buck" <trentbuck(a)gmail.com> Sent: Mon, 21/11/2011 11:37 To: "Christopher M. Bailey" <chris(a)chrisbailey.au.com> Cc: luv-main(a)luv.asn.au Subject: Re: [luv-main] DDNS Server [Scanned] Christopher M. Bailey wrote: > Hi All, > > I'm looking for a DDNS package that I can setup on a CentOS 5 server for potentially >thousands of devices deployed at thousands of sites world wide. I have had a quick look at >Bind9, which I think may buckle under this load if deployed by itself. I have also looked >gnudip, but would like some feedback on how easy it is to setup and maintain, also any other >packages people have used. > >Normally dynamic DNS works something like this: > > curl http://dyndns.org/update.cgi -F domain=foo.example.net -F password=UNPRINTABLE -F >address=1.2.3.4 > >Since you're talking about bind as a solution, I guess you're >basically talking about implementing an alternative to dyndns.org >(i.e. the hub) rather than a dyndns client (i.e. the spokes) ? > >In that case, why are you talking about bind? You're not thinking of >clients pushing updates to the hub via NOTIFY + IXFR, are you? >Because I don't see how that could possibly work. > > This is exactly what I'm trying to do and why I asked the question here. If Bind is not the right tool, what then is? TIA, again Chris -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.

1 0