I have pruned log via the php4log configuration file. If you remove info log entries will drop to a trickle.
Sent from my mobile. On Sep 3, 2013 7:38 AM, "José Juan Montes" <[email protected]> wrote: > > > > Hi, >> I'm trying to install Scalr and I read in the wiki some requirements: >> > > Hello :). > > >> 1x m1.large for mysql >>> 1x m1.large for the web portal >>> 1x c1.medium for cron jobs >> >> >> Is all these instances really required? why is it so important to >> separate cron jobs? >> >> > I haven't gone production yet but I have been deploying and testing to a > "m1.small" the complete system, including a local MySQL database. Only a > bunch of servers (8-10) are being managed, and the truth is the host is > pretty loaded permanently (~5). However, it runs just fine. I will scale it > to a bigger machine to get the load down to a reasonable value, but my > recommendation is to use a m1.small for the whole thing if you just want to > start deploying it. > > From my current observations, you definitely don't need such a big > instance if you plan to administer a few dozens or hundreds of servers. I > plan move the database to Amazon RDS anyway. It is always good to separate > database servers from applications, but you can do this afterwards and > definitely you can run MySQL locally even in a "small" instance for a start > if you wish. > > Regarding cron jobs, I believe it's important to ensure run cron jobs in > only one instance. So, if you had several Scalr nodes, you should run cron > jobs only in one of them, otherwise you'd possibly experience problems. > Given this, if your scalr node is loaded, the first thing to consider is to > move cron jobs away. But again, I have tested this on a single node. > > >> And, most of all, how much space I have to set up for mysql? What is your >> experience about it? >> > > I'm still running in 8GB (m1.small instance). After 15 days with ~10 hosts > the database is 1GB. 99% are log entries. Log data grows heavily and needs > to be pruned, not sure if this happens automatically at some point. Also, > as a reference, here's the size of my zipped backups. Note that most of it > is logging anyway. > > root@scalr01:/var/backups/mysql# ls -sh1 > total 232M > 26M scalr-20130829-063141.sql.gz > 37M scalr-20130830-064455.sql.gz > 47M scalr-20130831-062700.sql.gz > 57M scalr-20130901-063506.sql.gz > 68M scalr-20130902-064331.sql.gz > > You will also need some disk space for metrics data (stored in RRD files). > Around 5 files are stored per host, plus some others per role and per farm. > Currently I'm using 38 MB for that which is pretty low and affordable. I > don't expect this to be an issue, but we could easily move this directory > to a separate EBS volume if needed. > > Best! > > -- > You received this message because you are subscribed to the Google Groups > "scalr-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/groups/opt_out. > -- You received this message because you are subscribed to the Google Groups "scalr-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
