I'm trying to manage a process with supervisord, but the process does not have an option to run in foreground: it always daemonizes. (That's Zabbix Server).
Is there any way to manage daemons with supervisor? Any tools which will make it run in foreground? Or maybe, use the pidfile somehow?
I'm deploying a 3rd-party application in compliance with the 12 factor advisory, and one of the points tell that application logs should be printed to stdout/stderr: then clustering software can collect it.
However, the application can only write to files or syslog. How do I print these logs instead?
I have a wildcard certificate for *.example.com (signed by GeoTrust) and am going to host lots of application instances on different servers, each on my subdomain.
I am not willing to copy my OpenSSL private key to every customer. How do I use my certificate to create sub-certificates for individual subdomains without sharing mine?
In Linux, I have the following id:
uid=1005(username) gid=1005(username) groups=1005(username),33(www-data),1002(git)
I'd like to change my effective gid to git so that everything I create will belong to gid group. E.g. touch testfile gives owner=username group=git
How?
Is it always true that when a new file is created in a directory - the directory's mtime changes?
I'd like to recursively chmod a directory so that:
How to do it better, shorter, fancier? :) Maybe, use umask somehow?
All find solutions are too long: I always end with Copy-Paste :)
Looks like my USB HDD is old and corrupted:
rsync: read errors mapping "path/to/file.jpg": Input/output error (5)
Syslog:
kernel: [58927.940676] sd 6:0:0:0: [sdb] Unhandled sense code
kernel: [58927.940683] sd 6:0:0:0: [sdb] Result: hostbyte=invalid driverbyte=DRIVER_SENSE
kernel: [58927.940690] sd 6:0:0:0: [sdb] Sense Key : Medium Error [current]
kernel: [58927.940696] sd 6:0:0:0: [sdb] Add. Sense: Unrecovered read error
kernel: [58927.940703] sd 6:0:0:0: [sdb] CDB: Read(10): 28 00 00 21 f2 df 00 00 08 00
kernel: [58927.940716] end_request: critical target error, dev sdb, sector 2224863
The strange thing is that FS is not spoilt: I can list all directories. However, reading the contents of about 10% files gives these errors. Adiitionally, most of small files are safe, however, large files are all spoilt (corruption threshold is ~5MB).
How can it be? What options do I have to investigate & rescue the files? Can it be a driver issue? I can't believe that HDD damage could miss the FS metadata!
P.S. this HDD is perfidious: I have always been doing backups, and it waited for the right moment to betray :-D
SSH auth by RSA-key: ~/.ssh/authorized_keys. Okay.
Is there a way to set a system-wide authorized key so I can login to any account, including the newly created ones?
The reason: i've got a server and am the only admin. User accounts are used to isolate web-sites from each other just in case any of them is hacked.
I use SFTP and create new accounts often, thus, i'm tired of copying ~/.ssh/authorized_keys :)
How do I implement ctrl+c handling in bash scripts so that the script is interrupted, as well as the currently running command launched by the script?
(Imagine there's a script that executes some long-running command. The user hits ctrl+c and interrupts the command, but the script proceeds.) I need it to behave in a way that they are both killed.
My dnsmasq is configured:
All '*.lan' hosts registered via DHCP do resolve fine: dnsmasq known which IPs are assigned to which hostnames. However, 'rtfm.lan' is read from /etc/hosts and resolve to 127.0.0.1.
I can prevent dnsmasq from reading /etc/hosts and add an address manually:
no-hosts
address=/rtfm.lan/192.168.1.2
But this is not flexible: if one day I decide to change my IP — i'll also have to change it in dnsmasq.conf.
How can I tell dnsmasq to use local-hostname with its real IP?
Just in case.. here's my DNSmasq config:
# Never forward plain names (without a dot or domain part)
domain-needed
# Never forward addresses in the non-routed address spaces.
bogus-priv
# Read resolv.conf serially
strict-order
#==========[ NAMESERVER ]==========#
# Cache size
cache-size=4096
# Don't read /etc/hosts
no-hosts
# Read additional hosts-file (not only /etc/hosts) to add entries into DNS
addn-hosts=/etc/hosts-dnsmasq
# Auto-append <domain> to simple entries in hosts-file
expand-hosts
#=== HOSTNAME OVERRIDES
address=/localhost/127.0.0.1 # *.localhost => 127.0.0.1
#==========[ DHCP ]==========#
# Enable for the local network?
dhcp-authoritative
# Tell MS Windows to release a lease on shutdown
dhcp-option=vendor:MSFT,2,1i
#=== DHCP
# Domain name
domain=lan
# DNS-resolve hosts in these domains ONLY from /etc/hosts && DHCP leases
local=/lan/
# DHCP range & lease time
dhcp-range=192.168.1.70,192.168.1.89,24h
# Default route
dhcp-option=3,192.168.1.1
#=== FIXED LEASES
# LAN MY HOSTS
dhcp-host=00:23:54:5d:27:fa, rtfm.lan, 192.168.1.2
dhcp-host=00:23:54:5d:27:fb, rtfm.lan, 192.168.1.2
dhcp-host=c8:0a:a9:45:f1:03, 00:1e:64:9e:e9:5e, wtf.lan, 192.168.1.3
I'm going to move a small office to Ubuntu. Typical tasks: VoIP, E-Mail, documents, RDP access, ERP, printing, file-sharing.
What stack of technologies & solutions can you suggest to use for better experience and flexibility? Real-life success stories are welcome :)
Here's a list of possible features to implement:
I've launched something that took lots of memory and now everything lags a lot. I guess all applications' memory has gone to swap in order to free some space for the memory-intensive process, and now everything is slowly returning to RAM when accessed.
Is there a way to explicitly move everything possible from swap back to RAM? Or maybe not everything, but just some particular processes data?
I have MySQL data files on a thumb drive that are used on two hosts where mysql user's UIDs differ. As a result, MySQL fails to start when it's files have 0700 perms and an unknown UID as an owner.
I failed to find how to change MySQL's umask (and actually I don't like the idea of sharing these files to everyone), therefore I want to change UID of mysql user on both hosts so the files belong to the same user.
I'm going to change the UID and to chown all files owned by old mysql UID to the new user:
usermod --uid 900 --gid 900 mysql # assign the new uid
olduid=67 find / -user $olduid -group $olduid -print0 | xargs -0 chown "mysql:mysql"
Is this sufficient for an application to work in general case? Maybe, I have better options?
Are there any practical benefits in using rsyncd compared to rsync over ssh? Does it really increase speed, stability, anything?
I often use ImageMagick's convert for *->PNG conversion, but when PDF has more than 50 pages — convert eats more that 3 Gib (!!!) of memory. I guess it first loads everything.
That's unacceptable. It should read PDF page by page, why the heck all of them at once!
Maybe there's a way to tune it somehow? Or any good alternatives?
My client needs to have a virtualized Linux running on his Windows Server 2008 (R1). I've managed to setup VirtualBox as a service with srvany, but I see no way of creating a kind of 'stop script' so the virtual machine powers off correctly when I stop the registered service.
I know there may be other options of creating a headless VM, but that's not the point. How can I create a 'stop script' that is executed when I stop the registered service (or the system is rebooted)?
Or maybe, there are some other ways to launch 'system shutdown' scripts so VM is stopped properly on shutdown?
(for those that might be interested)
srvany: sc create UbuntuD binPath= "C:\Program Files (x86)\Windows Resource Kits\Tools\srvany.exe" start= auto DisplayName= "Ubuntu Daemon"[HKEY\_LOCAL\_MACHINE\SYSTEM\CurrentControlSet\Services\UbuntuD\Parameters] Application="C:\Program Files\Sun\VirtualBox\VBoxHeadless.exe" AppParameters="--vrdp=off -s Ubuntu" AppDirectory="C:\Program Files\Sun\VirtualBox\"
C:\Program Files\Sun\VirtualBox\vboxmanage, params: controlvm Ubuntu savestate
I'm going to build a high-performance web service. It should use a database (or any other storage system), some processing language (either scripting or not), and a web-server daemon. The system should be distributed to a large amount of servers so the service runs fast and reliable.
It should replicate data to achieve reliability and at the same time it must provide distributed computing features in order to process large amounts of data (primarily, queries on large databases that won't survive being executed on a single server with a suitable level of responsiveness). Caching techniques are out of the subject.
Which cluster/cloud solutions I should take for the consideration?
There are plenty of Single-System-Image (SSI), clustering file systems (can be a part of the design), projects like Hadoop, BigTable clones, and many others. Each has its pros and cons, and "about" page always says the solution is great :) If you've tried to deploy something that addresses the subject - share your experience!
UPD: It's not a file hosting and not a game, but something rather interactive. You can take ServerFault as an example of a web-service: small pieces of data, semi-static content, intensive database operations.
For those who might be interested:
Related questions:
I've occasionally lost my config file "/etc/mysql/my.cnf", and want to restore it. The file belongs to package mysql-common which is needed for some vital functionality so I can't just purge && install it: the dependencies would be also uninstalled (or if I can ignore them temporarily, they won't be working).
Is there a way to restore the config file from a package without un-ar-ing the package file?
dpkg-reconfigure mysql-common did not restore it.
What GUI tools do you advise to improve the usability of Linux for those accustomed to doing system administration the windows-way?
Users new to Linux often get confused when they need to configure something: editing a text configuration file may seem hard to them, and they want a GUI to tick checkboxes and just get what they want. A short list includes: hardware issues, drivers, Xorg (especially!), performance, network settings, sharing, user accounts, etc.
P.S. I've looked through questions already asked here, and found that there's no real centralized list of applications. Let's unite our knowledge! Community Wiki ;)
Is there a way to make bash display stderr messages in red color?