Tag Info

New answers tagged

0

As usual, problem was hiding in another place of the system. I had this rule for iptables which I just copied from some manual: -A POSTROUTING -j MASQUERADE This rule was set to allow NAT for accessing clients to Internet. This rule wasn't limited with any conditions, so requests from localhost to localhost was NAT'ed and applications was detecting wrong ...


0

You can restore the files from your latest restore point using http://nicbedford.co.uk/software/systemrestoreexplorer/ Just copy them to your mysql folder (I would backup the old files before). Here you can see, how to restore the data from a backup: http://stackoverflow.com/questions/10680261/restore-mysql-database-folder-from-a-recovered-hard-disk In ...


0

Does your SQL server config allow access from the host/IP that is your linux server? If you can connect fine with workbench (presumably from another PC/host/network, then you need to look at your user permissions and where you are allowed to connect from. Check this page as it gives the config for setting/checking the host part.


0

What do you see in error log? /var/log/mysql/error.log It would be easier. It looks like your instance crashes, but in the error log there will be more details about why.


0

This is not that bad, and 400 updates /month is literally nothing. You are doing it fine.


0

Sadly, MySQL historically uses localhost as a socket based connection, instead of ip. You will not see 127.0.0.1 or localhost until you don't connect by socket. (Yes it's wrong)


2

If you run a SHOW MASTER STATUS\G you'll see something like this: mysql> show master status\G *************************** 1. row *************************** File: mysql-bin.000299 Position: 780437462 Binlog_Do_DB: Binlog_Ignore_DB: Executed_Gtid_Set: 075d81d6-8d7f-11e3-9d88-b4b52f517ce4:1-616637650, ...


0

Upgrading MySQL Workbench to version mysql-workbench-community.x86_64 6.3.3-1.fc20 solved the issue.


0

You can use logrotate for that. A sample logrotate configfile could look something like this, assuming that you use a cronjob to save a daily mysql-dump in /home/myuser/mysqldumps/mydb.sql: /home/myuser/mysqldumps/mydb.sql { daily missingok rotate 31 notifempty } Adding compress can be considered to reduce diskspace.


2

This is no issue. First of all, SSDs have greatly improved during the last years. Overprovisioning and wear levelling (and to a small amount, the TRIM command, though not applicable in your case) have made them quite suitable as heavy-duty, general-purpose disks. I am not using anything but SSD on my development PC (which regularly does a lot of compiling) ...


0

the problem is with variable inside single quote: sudo debconf-set-selections <<< 'mysql-server-5.6 mysql-server/root_password password $mysql_pass' sudo debconf-set-selections <<< 'mysql-server-5.6 mysql-server/root_password_again password $mysql_pass_again' When I change it to double qoutes "" it is working...


4

Let's assume your import involves no updates and no deletions. So you are doing all insertions. This should only be writing new data to the transaction log. This means as data is added, it is always being written to a new sector. There might be some buffers/swap that gets churned/written to multiple times, but ignoring that, all of those inserts would ...


1

I helped him find a workaround in chat: Evidently his MySQL Workbench app is not able to properly set up the SSH forwarding tunnel, so I had him set it up manually using the OpenSSH ssh client as follows: ssh -L 33000:192.168.x.x:3306 user@host Then, once that is running, he was able to do a direct connect in MySQL Workbench to localhost:33000 and it ...


-2

As the poster of this writeup on SSDs said, what is really harmful is again and again writing small chunks of data. bits are stored into {1,2,3}-bit cells. These have limited lifespan. cells are grouped into [2-16]KB pages (smallest writeable unit) pages are grouped into (128-256 page-)blocks (smallest eraseable unit) for a page to be rewritten, it---and ...


1

If you are truly interested in figuring out the details then you will need the following question answered: On average how many bytes are in each row? If you can tell me that there are 10 columns, each column is varchar(100), and the encoding is UTF-8 then I can guess at worst case scenario that you have 4,000 bytes worth of data per row and add some more ...


17

Reads are fine, and SSD's can have their bits read from without any detrimental effect. Writes are another matter. Clearing a bit affects the integrity of the bit and after a lot of sequential writes, the bit will stop accepting new writes altogether. It can however still be read. Let me just say that the write limits on new enterprise drives are huge. ...


24

It really is not a straightforward answer to this. SSDs do not care about continuous writes as much as how many times any particular sector is overwritten. When SSDs first came out, something like SQL was a bad word as the operating system in general treated the drive like a traditional HDD and failures were very frequent. Since then, drives have become ...


12

Writing to SSDs isn't necessarily bad. It's the writing and rewriting of a single block that's bad. Meaning if you write a file delete it then write it again, or make small amounts of changes to a file over and over again. This causes wear on the SSD's. Databases would definitely fit into this category. However according to this article, petabytes of ...


1

SSD's do not like it. If you keep max write speed up for 5-10 years (24 hours per day, 7 days per week) then you might end up with a broken SSD. Ofc. After 5 years most servers have reached their economical end of life. Disclaimer: Do not try this with the very first generation of SSD. Those where less robust.


0

Regarding synchronizing your home directory with Unison and excluding certain system directories, there are two options. The first is to explicity list everything you want to sync and leave out any system-specific directories: # unison-profile.prf root = /home/user root = ssh://user@VM//home/user path = Documents path = .bashrc . . . The other option ...


2

Apparently, the server logging was disabled. I had to edit /etc/my.cnf.d/server.cnf and add the following line under section mysqld: log-error=/var/log/mysqllog After that, I could get proper logs which then allowed me to determine the problem (in my case it was an access right issue to the folders after a restore.)


1

I'll start by mentioning my own question posted on stack-overflow about a year ago, and it's somewhat of the same question: How can I set client configuration for MySQL Workbench without an installed server? As already mentioned in an answer here, as of MySQL WB 6 you can set the "leave autocommit mode enabled by default" preference in Preferences > SQL ...


0

First, confirm that skip-networking is commented out in your my.cnf. You will also need to allow remote connections from your IP to your database. GRANT ALL ON yourdatabase.* TO mysqluser@'yourip' IDENTIFIED BY 'yourpassword'; Lastly, confirm that your firewall is not blocking the connection.


1

Make sure you're using MySQL, since AUTO_INCREMENT doesn't work for other DBs such as SQL Server (use Identity(1, 1) instead with it). Use Invoice_No and not Invoice No when marking it as the PK.


1

It's likely less about the traffic that it is creating and more about the server load it's creating. Without seeing the code, I can't be of much help but I'd look into how often you are polling and things like that. Try putting it in a VM and seeing what kind of performance impact it's having on the system.


0

I believe the culprit is the fact that you in your first <VirtualHost> you have a ServerAlias *.abc.biz. Looks like that one covers any subdomains as well. Here's a snippet from my apache config, and it works for me. This is what I do for subdomains, and I have a lot of them: <VirtualHost *:80> ServerName example.com ...


0

/usr/local/mysql links to /home/ec2-user/install/mysql-5.6.24-linux-glibc2.5-x86_64, and while the link itself has the default link permission, we still do not know what the permissions are for /home/ec2-user/install/mysql-5.6.24-linux-glibc2.5-x86_64, and I bet that is the source of the issue. What is the output of ls -la ...


0

In my case: ls -ld /usr/local/mysql* lrwxr-xr-x 1 root wheel 27 Apr 10 16:43 /usr/local/mysql -> mysql-5.6.24-osx10.8-x86_64 drwxr-xr-x 17 root wheel 578 Apr 10 16:43 /usr/local/mysql-5.6.24-osx10.8-x86_64 So I set the new path into my PATH environment: cd $HOME cat .bashrc export PATH="$PATH:/usr/local/mysql/bin/" source .bashrc ...


0

The Query window (which can only be used for SELECT statements) by default has a Base parser which is optimized for the most common functions in an old version of HSQLDB (version 1.8 is what is bundled with OpenOffice, while the current version is 2.3). If you need to use a syntax in the query window that the parser objects to, you can turn off the parser ...


0

I had the exact same issue after restoring from a Time Machine backup. I had to completely reinstall MySQL to fix it as my /usr/local/mysql/data/mysql directory was corrupted. I followed these steps: Moved com.mysql.mysql.plist out of /Library/LaunchDaemons Moved my old mysql installation out of /usr/local into /tmp Downloaded latest version of MySQL ...


0

I think that I have found the solution. I have been running the ndb_restore via a script which I have been executing through a SSH connection as I do not currently have physical access to the servers. As such, when the SSH session dies, ndb_restore received a SIGHUP. I believe that this is what has been causing the process to stop without any notice or log ...



Top 50 recent answers are included