Lennart's weblog

Notes about open source software, computers, other stuff.

Author: LCK (Page 7 of 14)

A bioinfomatician on the move...

Exit a Bash script if an error occurs

5 July 2013 / / 0 Comments

Last week I found out that a Bash script I wrote to do some data QC gave me a false sense of security: a script continues even if one (or more) of the statements in the scripts fails (with an exit status not equal to 0). It turned out that for some of the data sets the QC wasn’t done correctly because I didn’t check the exit status after each step.

My first thought was: oh boy, that means I have to check $? for every step. That means a lot of repetitive code to write! Luckily my colleague came with the answer: add

set -e

at the top of you Bash script and the script will fail if one of its statements fails (for the fine print see the top answer in this is StackOverflow post).

Related Images:

Speeding up grep when looking in large files

13 June 2013 / / 0 Comments

In my line of work it is not uncommon to have to find out whether a given term is present in a long list. Say, for example you need to look up whether a set of, say 10, SNPs is present in a (possibly annotated) list of SNPs present on a genotyping array (having for example 240k SNPs).
My first instinct in such cases is to use grep, and it’s a good instinct that has served me well over the years.

Recently we had a case that involved quite some larger files. We needed to see whether a set of genomic positions was present in a genome-wide list of such positions. Of course we split the files up per chromosome, but still this took ~ 24 hours for a chromosome when using

grep -w -f short_list long_file > results

I was convinced this could be done faster and googled a bit, read the grep man page to find out that the -F option of grep ensures that the search string is not seen as a (regexp) pattern, but as fixed. This meant an enormous speed improvement. Instead of having to wait for 24 hours we got the output in under a minute!

I did a quick performance comparison: looking up ten items in a ~415MB file with 247,871 rows and 136 columns took ~2 minutes, 3 seconds with out -F and less than a second with the -F option:

$ time grep -w -f shortlist.txt largefile.tsv > out_withoutF
 
real    2m3.181s
user    2m0.780s
sys     0m2.196s
$ time grep -wF -f shortlist.txt largefile.tsv > out_withF
 
real    0m0.568s
user    0m0.500s
sys     0m0.060s

Related Images:

Fixing the NFS check plugin in Nagios (in Ubuntu)

11 June 2013 / / 0 Comments

For some time (probably after an upgrade, I actually don’t remember anymore) we had problems with the NFS check in Nagios on our Ubuntu 12.04 servers. The check would return UNKNOWN: RPC program nfs udp is not running. When running the actual check from the command line:

/usr/lib/nagios/plugins/check_rpc -H '$HOSTADDRESS$' -C nfs -c2,3

the output would be: Can't fork for rpcinfo.
It turns out that the file /usr/lib/nagios/plugins/utils.pm has the wrong path to the rpcinfo binary. Instead of /usr/sbin/rpcinfo it lists /usr/bin/rpcinfo. So, like most of the times, the fix is easy, but pinpointing the exact problem isn’t.

Don’t forget to restart Nagios after changing the path as utils.pm needs to be reloaded.

As Ubuntu is based on Debian, I expect this fix to work there as well. According to this Launchpad bug report this issue was fixed in January in version 1.4.16-1ubuntu1 of the nagios-plugins package, which is not in Ubuntu 12.04.

Related Images:

Importing a git repo into another one (keeping all history)

6 June 2013 / / 0 Comments

Today I wanted to create a new git repository that should contain several subdirectories that each were initially stored as separate git repos. Of course I didn’t want to lose the history. Thanks to user ebneter‘s answer at StackOverflow I was able to do so. These are the steps I took:

mkdir new_combined_repo
git init                           # Make empty new 'container' repo (no need to create a subdir at this point yet)
git remote add oldrepo /path/to/oldrepo
git fetch oldrepo
git checkout -b olddir oldrepo/master
mkdir olddir
git mv stuff olddir/stuff          # as necessary
git commit -m "Moved stuff to olddir"
git checkout master                
git merge olddir                   # should add olddir/ to master
git commit
git remote rm oldrepo
git branch -d olddir               # to get rid of the extra branch before pushing
git push                           # if you have a remote, that is

Related Images:

Converting from bzr to git

6 June 2013 / / 0 Comments

I’m in the process of moving several of my projects that used Bazaar (bzr) for revision control to Git. Converting a repository from bzr to git is very easy when using the fastimport package. In a Debian-based distribution run the following command to install the package (don’t be fooled by its name, it also contains the fastexport option):

sudo aptitude install bzr-fastimport

The go into the directory that contains your bzr repo and run:

git init
bzr fast-export `pwd` | git fast-import

You can now check a few things, e.g. running git log to see whether the change log was imported correctly. This is also the moment to move the content of your .bzrignore file to a .gitignore file.

If all is well, let’s clean up:

rm -r .bzr 
git reset HEAD

Thanks to Ron DuPlain for his post here, from which I got most of this info.

Related Images:

Solving “RTNETLINK answers: File exists” when running ifup

28 March 2013 / / 11 Comments

On a server with multiple network cards I tried to configure the eth3 interface by editing /etc/network/interfaces (this was an Ubuntu 12.04 machine).

This was the contents of /etc/networking/interfaces:

# The loopback network interface
auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
        address xxx.yyy.zzz.mmm
        netmask 255.255.255.0
        gateway xxx.yyy.zzz.1
        dns-nameservers xxx.yyy.zzz.aaa xxx.yyy.zzz.bbb
        dns-search mydomain.nl

auto eth3
iface eth3 inet static
        address 192.168.4.1
        netmask 255.255.255.0
        gateway 192.168.4.1

When I tried to bring the interface up I got an error message:

$ ifup eth3
RTNETLINK answers: File exists
Failed to bring up eth3.

It took me a while to figure it out, but the problem was the gw line in the eth3 entry. Of course you can only have one default gateway in your setup. I missed this because I was also trying to add routes to networks behind the machine on the other end of eth3.
In the end, removing the gw line in the eth3 entry solved the problem.

My final /etc/networking/interfaces looks like this:

# The loopback network interface
auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
        address xxx.yyy.zzz.mmm
        netmask 255.255.255.0
        gateway xxx.yyy.zzz.1
        dns-nameservers xxx.yyy.zzz.aaa xxx.yyy.zzz.bbb
        dns-search mydomain.nl

auto eth3
iface eth3 inet static
        address 192.168.4.1
        netmask 255.255.255.0
        post-up /sbin/route add -net 192.168.1.0 netmask 255.255.255.0 gw 192.168.4.250
        post-up /sbin/route add -net 192.168.2.0 netmask 255.255.255.0 gw 192.168.4.250
        post-up /sbin/route add -net 192.168.3.0 netmask 255.255.255.0 gw 192.168.4.250
        post-down /sbin/route del -net 192.168.1.0 netmask 255.255.255.0
        post-down /sbin/route del -net 192.168.2.0 netmask 255.255.255.0
        post-down /sbin/route del -net 192.168.3.0 netmask 255.255.255.0

Update 2013-08-19: Removed network entries as per Ville’s suggestion.

Related Images:

Pairing a device with a Logitech unifying receiver in Linux

8 February 2013 / / 2 Comments

My girlfriend’s keyboard and mouse stopped working some time ago. It turned out that her Logitech unifying receiver (a small USB dongle for keyboard and mouse) was a bit broken, only when twisted in a certain way it would work. So, I called Logitech, explained the situation and they offered to send us a replacement for free. Well done Logitech support!

Now, since we both use Linux as our main OS, the question was how to pair the mouse and keyboard with the new receiver. Logitech provides a piece of Windows software, but nothing for Linux. It turns out it’s not that difficult and you can find various little C programmes that do it for you. I tried Travis Reeder’s solution and it worked like a charm on my Ubuntu 12.04 machine.

These are the steps I took.
First I switched off the keybord and the mouse, then ran the following:

$ git clone https://github.com/treeder/logitech_unifier.git
Cloning into 'logitech_unifier'...
remote: Counting objects: 35, done.
remote: Compressing objects: 100% (26/26), done.
remote: Total 35 (delta 11), reused 33 (delta 9)
Unpacking objects: 100% (35/35), done.
$ cd logitech_unifier/
$ ./autopair.sh 
Logitech Unified Reciever unify binary not compiled, attemping compilation
Logitech Unified Reciever unify binary was successfully compiled
Auto-discovering Logitech Unified Reciever
Logitech Unified Reciever found on /dev/hidraw0!
Turn off the device you wish to pair and then press enter
[sudo] password for lennart: 
The receiver is ready to pair a new device.
Switch your device on to pair it.

I ran the autopair.sh script twice, once for the mouse and once for the keyboard.

Thanks Travis!

Related Images:

Using CUA selection mode in Emacs to edit rectangles

23 January 2013 / / 0 Comments

Today Planet Emacsen brought me Irreal’s second blog post in a short time on CUA mode in Emacs. So far I’ve always ignored it because as far as I knew CUA mode is about getting the Windows keyboard shortcuts of Ctrl-c, Ctrl-x and Ctrl-v for copying and pasting to Emacs. The thing is, I date back to the DOS era when Shift-Del and Shift-Ins were used for that, so back in my Windows days I never got used to those ‘new’ keyboard shortcut. Now that I’ve been an Emacs user for more than a decade I’m so used to C-w and C-y and I see no reason for having the Windows shortcuts work in Emacs.

Back to Irreal. In his recent blog posts he writes about a subset of cua-mode: cua-selection-mode. The video by Mark Mansour that we writes about says it all (it’s short, so go and watch it!). What cua-selection-mode is all about is rectangle editting. So far I’ve been using the regular Emacs keys for rectangle selection and editing (basically C-space to select a rectangle and C-r-k to cut it, C-r-t to insert text and C-r-y to paste a rectangle). By setting

(cua-selection-mode 1)

in your ~/.emacs file you only enable the rectangle features of CUA mode.

So, for those that didn’t watch the video, what does the rectangle editing mean? It means that you can for example simply insert a list of increasing numbers in a text (this may come in handy in an org-mode table for example), or you can insert the same text in front of and/or behind a selected column of text.

Key combos to remember are:

  • C-return: Start selection
  • return: move the cursor to top-left, top-right, bottom-left and bottom-right corner of the selected rectangle
  • C-?: briefly list the available key combinations (with rectangle selection enabled)
  • M-i: if the selection is a column of numbers increase the numbers (by one)
  • M-n: Insert a number in the column (asks for start value and increment value)
  • C-1 C-w: Kill (cut) the contents of the rectangle to register 1 (you can use number 0–9 for different registers). Using C-1 C-y yanks (pastes) the rectangle at the cursor position.

Related Images:

Comparing rsnapshot and obnam for scheduled large backups

4 January 2013 / / 4 Comments

Introduction

The home directories of the servers I administer at work total about 6.5TB of data. The home directories are stored on a file server (using ext4 partitions) and served to the other server over NFSv3 with a bonded 1Gbps LAN link.

As you all know backups are a good idea but how to implement a backup strategy for this kind of data? We decided quite early that using tapes as backup medium was out of the question. We simply can’t afford them given the amount of disk space we need. Moreover, tapes usually require operator involvement and neither me nor my colleague feels like going to the data centre every week. Our idea was to back up to another server with enough disk space in a different part of the data centre. For off-site backups we can always make an annual (maybe monthly) backup either on tape at SurfSARA/BigGrid or on a remote server.

Before implementing a given strategy several things need to be known and tested. The major questions we wanted to have an answer to were:

  1. How often do we want to backup the data? Daily snapshots? Weekly? Monthly?
  2. How many of the backups mentioned above do we want to keep? And for how long?
  3. In order to answer these questions (given a roughly fixed amount of backup space) we need to know
    • How much data changes per night/week/etc.
    • How much duplication is there in the data? How many people store the same file (or blocks, if you go for block-level deduplication)?
  4. Is NFS/network speed a limiting factor when running the backups?
  5. Can the tool preserve additional file system attributes like POSIX ACLS?

Candidates

After looking around the web and looking back at my own experiences I came up with three possible candidates. Each of them allows for backup rotation and preserves Posix ACLs (so points 1 and 5 above have been taken care of).

  1. Bacula: enterprise-level backup application that I’ve used in combination with tapes in the past. Easily supports multiple clients, tape robots, etc. No deduplication. All metadata etc. are stored in a (MySQL) database, so restoring takes some effort (and don’t forget to make a backup of the database as well!).
  2. rsnapshot: based on rsync, makes snapshots using hard links. Easy to restore, because files are simply copied to the backup medium.
  3. rdiff-backup: similar to rsnapshot, but doesn’t allow for removal of intermediate backups after a given time interval. Consequently it was the first candidate to fall of my list.
  4. Obnam: a young tool that promises block level data deduplication. Stores backed up data in its own file format. Tools for browsing those archives are not really well developed yet.

Tests

Because I already had quite some experience with Bacula but none with the other two candidates (although I use rsync a lot) I decided to start a test run with Obnam, followed by a run with rsnapshot. These are the results:

Obnam

After backing up /home completely (which took several days!), a new run, several days later took (timing by the Linux time command):

Backed up 3443706 files, uploaded 94.0 GiB in 127h48m49s at 214.2 KiB/s average speed830 files; 1.24 GiB (0 B/s)

real    7668m56.628s
user    4767m16.132s
sys     162m48.739s

From the obname log file:

2012-11-17 12:41:34 INFO VFS: baseurl=/home read=0 written=0
2012-11-21 23:09:36 INFO VFS: baseurl=/backups/backup_home read=2727031576964 written=150015706142
2012-11-21 23:09:36 INFO Backup performance statistics:
2012-11-21 23:09:36 INFO * files found: 3443706
2012-11-21 23:09:36 INFO * uploaded data: 100915247663 bytes (93.9846482715 GiB)
2012-11-21 23:09:36 INFO * duration: 460128.627629 s
2012-11-21 23:09:36 INFO * average speed: 214.179341663 KiB/s
2012-11-21 23:09:36 INFO Backup finished.
2012-11-21 23:09:36 INFO Obnam ends
2012-11-21 23:09:36 INFO obnam version 1.2 ends normally

So: ~5 days for backing up ~100 GB of changed data… Load was not high on the machines, neither in terms of CPU, nor in terms of RAM. Disk usage in /backups/backup_home was 5.7T, disk usage of /home was 6.6T, so there is some dedup, it seems.

rsnapshot

A full backup of /home to (according to the log file):

[27/Nov/2012:12:55:31] /usr/bin/rsnapshot daily: started
[27/Nov/2012:12:55:31] echo 17632 > /var/run/rsnapshot.pid
[27/Nov/2012:12:55:31] mkdir -m 0700 -p /backups/backup_home_rsnapshot/
[27/Nov/2012:12:55:31] mkdir -m 0755 -p /backups/backup_home_rsnapshot/daily.0/
[27/Nov/2012:12:55:31] /usr/bin/rsync -a --delete --numeric-ids --relative --delete-excluded /home /backups/backup_home_rsnapshot/daily.0/localhost/
[28/Nov/2012:23:16:16] touch /backups/backup_home_rsnapshot/daily.0/
[28/Nov/2012:23:16:16] rm -f /var/run/rsnapshot.pid
[28/Nov/2012:23:16:16] /usr/bin/rsnapshot daily: completed successfully

So: ~1.5 days for a full backup of 6.3TB. An incremental backup a
day later took:

[29/Nov/2012:13:10:21] /usr/bin/rsnapshot daily: started
[29/Nov/2012:13:10:21] echo 20359 > /var/run/rsnapshot.pid
[29/Nov/2012:13:10:21] mv /backups/backup_home_rsnapshot/daily.0/ /backups/backup_home_rsnapshot/daily.1/
[29/Nov/2012:13:10:21] mkdir -m 0755 -p /backups/backup_home_rsnapshot/daily.0/
[29/Nov/2012:13:10:21] /usr/bin/rsync -a –delete –numeric-ids –relative –delete-excluded –link-dest=/backups/backup_home_rsnapshot/daily.1/localhost/ /home /backups/backup_home_rsnapshot/daily.0/localhost/
[29/Nov/2012:13:25:09] touch /backups/backup_home_rsnapshot/daily.0/
[29/Nov/2012:13:25:09] rm -f /var/run/rsnapshot.pid
[29/Nov/2012:13:25:09] /usr/bin/rsnapshot daily: completed successfully

So: 15 minutes… and the changed data amounted to 21GB.

This gave me a clear winner: rsnapshot! Not only is it very fast, but given its simple way of storing data restoring a backup of any file is quickly done.

We now also have answers to our questions: Our daily changing volume is of the order of ~ 100GB, there isn’t much data that can be deduplicated. We also monitored the network usage and, depending on the server load it can be limiting, but since a daily differential backup takes only 15-30 minutes that isn’t a problem.
For a remote backup sever that was connected with a 100Mbps line we did see that the initial backup took a very long time. We should try to get a faster connection to that machine.

The future

The next challenge we face is how to back up some of the large data sets we have/produce. These include aligned BAM files of next-generation sequencing data, VCF files of the same data, results from genomic imputations (both as gzip-ed text files and as binary files in DatABEL format). This also totals several TB. Luckily these files usually don’t change on a daily basis.

Related Images:

Booting an Ubuntu server with a degraded software RAID array

4 January 2013 / / 0 Comments

My home server runs Ubuntu 12.04 with a software RAID 5 array and since a couple of days I’ve been getting e-mails from the SMART daemon warning me of uncorrectable errors on one of the drives. Today I took the time to take the failing drive out and check it with the tools from the manufacturer.

Because I didn’t want to run the risk of unplugging the wrong drive with the system on (and thus losing the whole RAID array) I shut the server down, removed the harddrive and started it again. The idea was that it would boot right back into the OS, but with a degraded RAID array. Unfortunately the server didn’t come up… After connecting a keyboard and monitor to it it turned out that the system was waiting with an initramfs prompt. From there I could check that the RAID array was indeed degraded, but functioning fine as I could manually mount all partitions.

Some Googling later I found out that by default Ubuntu doesn’t boot into a degraded software RAID array. This is to make sure you as administrator know something is wrong. A good idea for a laptop or PC, but not for a standalone server. The solution is the following:

  • From the initramfs prompt mount your original filesystems, for example in /mnt.
  • Use chroot /mnt to change root into your server’s hard disks.
  • In the file /etc/initramfs-tools/conf.d/mdadm add or change the line to 
    BOOT_DEGRADED=true
  • Then run 
    update-initramfs -u

    to regenerate the initial ramdisk.

  • Type exit to exit the chroot environment.
  • Unmount your file systems and reboot

Now your server should continue booting even though it has a degraded RAID array.

Links

Related Images:

« Older posts Newer posts »

Archives

© 2024 Lennart's weblog

Theme by Anders Norén — Up ↑