Zein’s place on the net

For some time now I have been flirting with the idea of getting myself a Network Attached Storage (NAS). I get excited about the idea of having all my important files and stuff on a single device that supports raid 1. I can then reinstall Ubuntu on my laptop as many times as I want without having to worry about moving data somewhere safe first. I will not have to give up my laptop to my wife because that is where we keep all our personal digital photos and videos and she really wants to send the latest snapshots of our daughter to her mother… NOW.

If you look at what is available on the market you can see that the cost of these devices are tremendous. The much cheaper alternative is to build something myself. I had a dual PIII 800MHz machine but I thought that was an over kill. I want this to consume the least possible amount of power and to run as cool as possible. When I found a old celeron machine at Uni I decided that I would put something together. I attached 3x 14GB IDE hard disks and a CD drive and had myself an ugly but functional machine.

I looked at FreeNAS which is based on monowall which in turn is a based on a minimal FreeBSD system. What is really cool about FreeNAS that it easily fits on a USB or CF disk and moves your machine towards being an appliance with all the benefits that include less heat and better reliability.

Being a bit paranoid I decided I would not accept anything less than ZFS for my data. I believe it is the only file system that detects silent data corruption. It also has a lot of hype surrounding it and I have being wanting to play around with it for a while. The current versions of FreeNAS do not have ZFS support, but when FreeBSD 7.0 comes out it will have native ZFS support. This means that FreeNAS will gain ZFS support at some stage. I, however do not want to wait.

Enter Nexenta, a distribution based on opensolaris with a GNU userland. I wasn’t completely at home, coming from a Linux background and I noticed that some of the command line tools support different command line options, but here is the best thing about Nexenta… It uses apt for its package management. I truly believe that apt is the best tool out there and is what drew me to Debian in the first place. It also has some nifty features like support for ZFS root file system and a tool called apt-clone that takes a ZFS snapshot of the system before major upgrades and allows you to revert if the upgrade goes wrong. All this without rebooting so no downtime. To summarize Nexenta is an interesting project and I will use it to build my NAS.

I installed Nexenta Alpha7 and created a zpool out of 2 of the disks changed the mount point to /mydata and shared it via NFS (ZFS has built in NFS/NFS4)

zpool create mydata mirror c0d1 c1d0
zfs set mountpoint=/mydata mydata
zfs set sharenfs=on mydata

After a while I realized that the Alpha7 version was not the latest and I wasn’t able to upgrade to the latest via apt! I installed Nexenta 1.0 RC3 and was able to import my old (mydata) ZFS pool. running zpool import will show all available pools that can be imported.

root@sun:~# zpool import
  pool: home
    id: 14253621109838762705
 state: ONLINE
status: The pool is formatted using an older on-disk version.
action: The pool can be imported using its name or numeric identifier, though
        some features will not be available without an explicit 'zpool upgrade'.
config:
 
        home        ONLINE
          c0d0s1    ONLINE
 
  pool: mydata
    id: 7749473723610951541
 state: ONLINE
status: The pool is formatted using an older on-disk version.
action: The pool can be imported using its name or numeric identifier, though
        some features will not be available without an explicit 'zpool upgrade'.
config:
 
        mydata      ONLINE
          mirror    ONLINE
            c0d1    ONLINE
            c1d0    ONLINE

I had to use -f (force) to import my previous pool

root@sun:~# zpool import mydata
cannot import 'mydata': pool may be in use from other system
use '-f' to import anyway
root@sun:~# zpool import -f mydata

I was then able to upgrade the ZFS pool

root@sun:~# zpool upgrade
This system is currently running ZFS pool version 10.
 
The following pools are out of date, and can be upgraded.  After being
upgraded, these pools will no longer be accessible by older software versions.
 
VER  POOL
---  ------------
 3   mydata
 
Use 'zpool upgrade -v' for a list of available versions and their associated
features.
root@sun:~# zpool upgrade mydata
This system is currently running ZFS pool version 10.
 
Successfully upgraded 'mydata' from version 3 to version 10

And finaly I get to see my pool up and running

root@sun:~# zpool list mydata
NAME     SIZE   USED  AVAIL    CAP  HEALTH  ALTROOT
mydata  12.6G  9.30G  3.33G    73%  ONLINE  -
root@sun:~# zfs list mydata
NAME     USED  AVAIL  REFER  MOUNTPOINT
mydata  9.30G  3.13G  9.30G  /mydata

I must say I am happy with my NAS. All I need is to get some large disks in there and I am good to go. I also look forward to get some experience with Nexenta and ZFS. I wonder if you can fit Nexenta onto a 128MB or 256MB CF card?

Is this a poor geeks dream or what? It is like I have been dreaming for a small low cost laptop for years and then all at once every one wants in on this market.

Acer is looking to enter the low cost laptop market with 8″ – 9″ designs. Gigabyte also announced that it will produce what seams to be a low cost laptop due in June. Linux is not mentioned in either case but given the success of the other low cost machines with Linux I would say it is reasonable to assume that there would be a Linux version of both laptops.

In other news, Everex plans to expand their cloudbook line with some nifty ideas.

I can’t say I saw that coming, but it happened!

If the cool (and cheap) Linux laptops were not enough, Shuttle is going to produce $200 Linux PCs. Sure the specs are not so good, but that looks like a good mythtv frontend to me. Or maybe a NAS that doesn’t have to live in the closet?

Different people like different things. Different people fantasize about buying different things. I fantasize about gadgets. My latest fantasies have been about the ASUS EeePC. The 4G model sells for $399, has a 7″ 800×640 screen, runs Linux and looks beautiful.

This is the sort of laptop that I have always dreamed of having yet could never afford. The only thing that is not perfect is the relatively low 800×640 resolution. At CES this year, ASUS have announced 8″ and 8.9″ versions of the EeePC. The 7″ and 8″ EeePCs are the same size and the screens have the same 800×640 resolution. I am hoping that the 8.9″ EeePC has a 1024×600 resolution. Interestingly ASUS EeePcs have been seen running splashtop (an instant on Linux based OS). Could things get crazier ?

Also of interest is the Everex cloudbook (also $399) based on the VIA nanobook design. It has been announced to start selling on the 25th of this month. The cloud book as a longer battery life and a 30GB HDD (vs the 4GB SSD on the EeePC). If after looking at the pictures, you are wondering where the touch pad is then take a look at this video

Anyway between them there should be a very cool laptop waiting to be bought when I visit the USA in July. If you are as interested as I am then suscribe to the RSS feeds of Cloudbooker and Eeeuser

I was just reading slashdot and I found this comment. I found it so funny I thought to share it

MCSE (Minesweeper Consultants, Solitaire Experts).

I am writing this post from SC07 in Reno NV, USA. I am presenting a poster titled The Sony PlayStation 3 and the NVIDIA 8800 GPU: Performance and Programmability Evaluation for Machine Learning. Big title, huh?

I am really happy to be here. This is my first conference and so far I am impressed. Yesterday I attended the High Performance Computing on GPUs with CUDA tutorial. It seams that this tutorial had the largest registration in the history of SC. It was really good (even though I already new alot of the stuff). They had a very good talk on performance optimization which I am of course interested in as I am working on the NVIDIA 8800 GPU.

I hope that I can create some good contacts and meet some new people.

UPDATE:

I got an email a few days ago that pointed out to me the fact that my SC07 poster is not available for download anywhere! I am remedying this by uploading it here

Email has always been a problem for me. Using web interfaces is a pain. Even the new JavaScript, single-page apps nowadays are bloated and slow. You can’t compare them to evolution, thunderbird or even outlook Most Email Service Providers (ESPs) do provide POP3, but POP3 doesn’t support folders and is not a good option if you would like to access your mail from multiple locations.

So what is the solution? IMAP or the Internet Message Access Protocol gives me all the features I need:

• Multiple clients simultaneously connected to the same mailbox

  I check my email from my Desktop, Laptop and PC at Uni.

• supports folders

  can’t live without them. I don’t believe in labels, and virtual folders. I want to physically partition my mail.

• Read/Write Support 2-way Synchronization

  I am not sure what this is really called but you know how POP3 supports downloading only? Well with IMAP you can upload emails to you mailbox. This also allows you to copy emails between folders or even accounts. It also allows you to do things like have your sent mail folder in your mailbox and have all your clients upload the sent mail there.

So what is the problem then? Most ESPs do not support IMAP. My old ESP (gawab.com) did, but due to the stability issues of gawab, I had to look elseware.

Now I didn’t get your hopes up just to let you down! Remember that $5/month VPS I have? It is time to get some more use out of it!

The plan is to have an IMAP server on my VPS, I don’t want to have the responsibility of running a my own SMTP server, just the IMAP server. I would also like to consolidate my email addresses into one mailbox. This seams like a simple plan.. a good one too, if I don’t say so my self

Solving the SMTP Issue

So I don’t want the responsibility of running a SMTP server so I will pass the buck to google. Google Apps provides email for your domain so you don’t have to worry about 99.99% availability and all that.

Installing and Configuring the IMAP Server

I decided on Dovecot for the IMAP server because of it’s low memory footprint as well as it’s stability and good reputation.
in ubuntu all I needed was to:

apt-get install dovecot-imapd

I wanted to set up IMAP over SSL for better security so I needed to create a SSL certificate

cd /etc/ssl
openssl genrsa 1024 > private/domain.key
chmod 400 private/domain.key
openssl req -new -x509 -nodes -sha1 -days 365 -key private/domain.key > certs/domain.cert

I new that I planned to use a passwd like password file for authentication so I created one right away:

cat >>/etc/dovecot/passwd < <EOF
user@domain:{plain}password
EOF

Then you need to configure dovecot by editing /etc/dovecot/dovecot.conf

protocols = imaps
 
## Logging
log_timestamp = "%Y-%m-%d %H:%M:%S "
 
## SSL settings
ssl_cert_file = /etc/ssl/certs/domain.cert
ssl_key_file = /etc/ssl/private/domain.key
 
## Login processes
login_user = dovecot
 
## Mailbox locations and namespaces
mail_location = maildir:/var/lib/mail/%d/%n
mail_extra_groups = mail
 
## Mail processes
first_valid_uid = 107
last_valid_uid = 107
 
## IMAP specific settings
protocol imap {
  login_greeting_capability = yes
}
 
## Authentication processes
auth_debug = yes
auth_debug_passwords = yes
 
auth default {
  mechanisms = plain
  user = root
  passdb passwd-file {
    # Path for passwd-file
    args = /etc/dovecot/passwd
  }
  userdb static {
    args = uid=dovecot gid=mail domain=domain home=/var/lib/mail/%d/%u
  } 
}
 
## Dictionary server settings
plugin {
}

apt-get install getmail4
mkdir /etc/getmail
chown -R dovecot:mail /etc/getmail

create a getmailrc file /etc/getmail/domain-user that describes the retrieval and delivery phases
It will look something like:

[retriever]
type = SimplePOP3SSLRetriever
server = pop.gmail.com
username = user@domain
port = 995
password = your_password
 
[destination]
type = Maildir
path = /var/lib/mail/domain/user/
 
[options]
delete = false
verbose = 0
delete_after = 0
read_all = false

test it with:

sudo -u dovecot getmail --getmaildir=/etc/getmail  --rcfile=/etc/getmail/domain-user

if all is OK then create a cron job to run say every 15 minutes

# Crontab for getmail. Download mail every 15 minutes
0,15,30,45 * * * *    dovecot getmail --getmaildir=/etc/getmail  --rcfile=/etc/getmail/domain1-user1
5,20,35,50 * * * *    dovecot getmail --getmaildir=/etc/getmail  --rcfile=/etc/getmail/domain2-user2

And there you have it! Just one nifty IMAP mailbox that contains all you email from all your accounts.

كل عام و أنتم بخير

Happy Eid

About 2 months ago I got myself a Virtual Private Server(VPS) from http://www.vpsvillage.com. I am paying $5 a month for a Xen virtual machine with 64MB of RAM and a 4GB hard disk. I have lots of plans for my VPS. The most essential is hosting my website (basically this blog plus a small rails app). I immediately installed a LAMP environment and had my blog up and running in no time.

Now I have a very small number of hits per day… So small that I am embarrassed to post that number here:). Suffice to say that it would be better to calculate per week or month Anyhow my site site kept dieing on me. The Out of Memory thread (OOM) kept killing off processes on my server.

Rather that tune apache I decided to install lighttpd and try and tune the rest of my system. So I did a fresh install of 64-bit ubuntu (actually I now feel that the 64-bit part was a mistake but I am not sure enough to go through the whole install process again). I will attempt to describe how I went through the process of tuning my 64MB VPS.

Tuning Tools

I installed munin as it produces lovely graphs. The graphing process it quite heavy so to lower its impact I removed all unnecessary plugins (I kept: cpu, load, memory and swap only) and I also modified the cron script so that the graphs and html files are updated every 15 minutes (munin updates every 5 minutes)

I am now able to look at nice graphs of my memory usage

you can see munin in progress here

Tune Linux

Next thing I did was to try and tune the virtual memory subsystem (ie /proc/sys/vm/*). I fist tried to lower the swappiness value but that did not have the result I expected! You see I thought that a lower swappiness value would decrease the systems tendency to swap caches to disk. It in fact decreases the tendency to swap anything to disk. I was getting OOM with swap only a quarter full.
What I did at the end was:

• leave swappiness at the default value of 60
• set vfs_cache_pressure to 200
• set max_map_count to 1000

No doubt further understanding of the VM subsystem would allow more efficient tuning, but that will have to wait for another day.

Tune Lighttpd

With the small load I have and the small memory footprint of lighty (less than 1MB), I can’t see much to be done here! I have setup php o run as a fast-cgi process. I have set max-procs to 1 in lighty with PHP_FCGI_CHILDREN=1. This means we will have 2 php-cgi processes, one is the parent monitoring process (PP) and the other will actually serve requests (SP). I might increase PHP_FCGI_CHILDREN to 2 in the future.

Tune php

I wanted to see what I could do to decrease the memory footprint for my php-cgi processes so I hit my site with:
ab -n 50 http://azein.com/ and fired up top (using M to sort by memory usage) and got

2163 www-data 16 0 77396 13m 3100 S 0.0 21.6 0:08.36 php-cgi
2162 www-data 23 0 69212 6132 3668 S 0.0 9.4 0:00.00 php-cgi


I then edited my php.ini file and modified the following values:

memory_limit = 16M
register_long_arrays = Off
;engine = On

I also commented out the loading of the mysqli, pdo and pdo_mysql modules as I don’t use them.

I tried again with ab -n 50 http://azein.com/and got:

2278 www-data 16 0 71096 12m 3016 S 11.3 20.2 0:09.36 php-cgi
2277 www-data 24 0 62660 3596 3528 S 0.0 5.5 0:00.00 php-cgi

so we have a small decrease in memory foot print.

Tune MySQL

1st thing to do before you even start to tune your parameters is to skip-innodb this by itself is supposed to decrease the memory foot print by 100MB.

The 1st parameter I will tune is the key_buffer. This describes the amount of memory used to cache the tables keys and indexes. The sum of my *.MYI files are about 80K so I will set the key_buffer size to 512KB with lots of room for expansion.

The 2nd set of parameters are the threads and concurrency group, now I plan to have no more than 2 php-cgi threads running at any time so it make no sense to have more than 2 mysqld threads running for now. I also set the max connections to 10 as I do not expect any more than that.

max_connections = 20
thread_concurrency = 2


The 3rd set of parameters are the query cache configuration group. I do not have many queries that I think need to be cached. I suppose the front page of my blog (about 10 posts) is all that will benefit from this.

query_cache_limit = 64K
query_cache_size = 512K

mysqld now consumes just under 7MB but I will take a look at my hit/miss stats in a few days to how the query cache parameters are doing.

Last thing I did was turn off binary logging and I do not plan to use it.

Final Comments

No mater what I did I couldn’t get my rails app running without the system coming to a halt! The rails app has a memory footprint of 30MB and I don’t have the energy to see if it possible to bring that down and what the minimal memory footprint of a rails app is!

I could upgrade my VPS to 128MB RAM, but I think that 64M should be sufficient and I will continue trying to make the most out of it.