New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Why do my two identical VM's have such diff RAM usage?
Why is my RAM usage so different between the MASTER (NS1) and SLAVE (NS2).
Is this normal? Both exist on two different VPS nodes and both are in separate cities. Same hosting provider.
Both running Ubuntu 12.04 LTS + MySQL server + PDNS (authoritive)
NS2 = SLAVE
NS1 = MASTER
root@ns1:~# free -m
total used free shared buffers cached
Mem: 768 364 403 0 0 28
-/+ buffers/cache: 336 431
Swap: 0 0 0
root@ns2:~# free -m
total used free shared buffers cached
Mem: 768 39 728 0 0 12
-/+ buffers/cache: 26 741
Swap: 0 0 0
Comments
Post your TOP
root@ns1:~# topInstalled services/programs.
@Maximum_VPS - they are IDENTICAL. Spawned at the same time, installed with the same programs. The only difference is one is the master and the other the slave in MySQL.
@Noerman
NS1
top - 05:40:25 up 13:39, 3 users, load average: 0.07, 0.02, 0.00 Tasks: 27 total, 1 running, 26 sleeping, 0 stopped, 0 zombie Cpu(s): 0.0%us, 0.0%sy, 0.0%ni,100.0%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 786432k total, 373752k used, 412680k free, 0k buffers Swap: 0k total, 0k used, 0k free, 80452k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 1 root 20 0 24144 1164 464 S 0 0.1 0:00.12 init 2 root 20 0 0 0 0 S 0 0.0 0:00.00 kthreadd/10966 3 root 20 0 0 0 0 S 0 0.0 0:00.00 khelper/10966 76 root 20 0 17188 312 204 S 0 0.0 0:00.00 upstart-udev-br 86 root 20 0 21528 352 320 S 0 0.0 0:00.00 udevd 110 root 20 0 49912 1092 900 S 0 0.1 0:00.07 sshd 207 messageb 20 0 23772 288 180 S 0 0.0 0:00.00 dbus-daemon 245 root 20 0 14924 364 360 S 0 0.0 0:00.00 xinetd 257 root 20 0 19068 440 328 S 0 0.1 0:00.03 cron 284 mysql 20 0 1339m 27m 3640 S 0 3.5 0:06.33 mysqld 329 syslog 20 0 12708 524 396 S 0 0.1 0:00.12 syslogd 399 root 20 0 21524 224 212 S 0 0.0 0:00.00 udevd 400 root 20 0 21524 244 212 S 0 0.0 0:00.00 udevd 474 root 20 0 15144 300 104 S 0 0.0 0:00.00 upstart-socket- 537 root 20 0 78620 924 220 S 0 0.1 0:00.00 saslauthd 538 root 20 0 78620 720 16 S 0 0.1 0:00.00 saslauthd 600 root 20 0 89044 2188 496 S 0 0.3 0:00.60 sendmail-mta 643 root 20 0 69728 1940 436 S 0 0.2 0:00.58 apache2 646 www-data 20 0 69800 2108 544 S 0 0.3 0:00.00 apache2 1589 root 20 0 73432 2052 1164 S 0 0.3 0:00.19 sshd 1603 root 20 0 18028 1260 768 S 0 0.2 0:00.00 bash 1627 root 20 0 73316 2016 1176 S 0 0.3 0:00.17 sshd 1641 root 20 0 18028 1256 768 S 0 0.2 0:00.00 bash 2952 www-data 20 0 69752 2084 552 S 0 0.3 0:00.00 apache2 3345 root 20 0 73432 3660 2764 S 0 0.5 0:00.02 sshd 3357 root 20 0 18028 1984 1492 S 0 0.3 0:00.00 bash 3740 root 20 0 17164 1324 1052 R 0 0.2 0:00.00 topNS2
top - 05:41:32 up 14:12, 2 users, load average: 0.02, 0.02, 0.04 Tasks: 27 total, 1 running, 26 sleeping, 0 stopped, 0 zombie Cpu(s): 0.0%us, 0.0%sy, 0.0%ni, 99.9%id, 0.0%wa, 0.0%hi, 0.0%si, 0.1%st Mem: 786432k total, 45664k used, 740768k free, 0k buffers Swap: 0k total, 0k used, 0k free, 18016k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 1 root 20 0 24148 788 364 S 0 0.1 0:00.19 init 2 root 20 0 0 0 0 S 0 0.0 0:00.00 kthreadd/11093 3 root 20 0 0 0 0 S 0 0.0 0:00.00 khelper/11093 106 root 20 0 17188 196 192 S 0 0.0 0:00.00 upstart-udev-br 117 root 20 0 21528 228 224 S 0 0.0 0:00.00 udevd 155 root 20 0 21524 192 188 S 0 0.0 0:00.00 udevd 156 root 20 0 21524 192 188 S 0 0.0 0:00.00 udevd 168 messageb 20 0 23772 180 176 S 0 0.0 0:00.01 dbus-daemon 196 root 20 0 15144 124 120 S 0 0.0 0:00.00 upstart-socket- 365 root 20 0 19068 308 236 S 0 0.0 0:00.05 cron 431 syslog 20 0 12708 464 332 S 0 0.1 0:00.48 syslogd 452 root 20 0 49912 492 380 S 0 0.1 0:00.20 sshd 485 root 20 0 78620 176 172 S 0 0.0 0:00.00 saslauthd 486 root 20 0 78620 20 16 S 0 0.0 0:00.00 saslauthd 647 root 20 0 14924 236 232 S 0 0.0 0:00.00 xinetd 672 root 20 0 89044 572 388 S 0 0.1 0:01.29 sendmail-mta 705 root 20 0 69728 300 260 S 0 0.0 0:01.12 apache2 708 www-data 20 0 69768 264 252 S 0 0.0 0:00.00 apache2 714 root 20 0 73316 1016 752 S 0 0.1 0:00.58 sshd 726 root 20 0 18028 224 220 S 0 0.0 0:00.03 bash 3650 mysql 20 0 1463m 2400 400 S 0 0.3 0:09.43 mysqld 4570 root 20 0 97956 436 372 S 0 0.1 0:02.16 pdns_server 4836 www-data 20 0 69752 240 228 S 0 0.0 0:00.00 apache2 4900 root 20 0 480m 640 420 S 0 0.1 0:00.02 pdns_server 7324 root 20 0 73432 3652 2764 S 0 0.5 0:00.04 sshd 7340 root 20 0 18028 1988 1492 S 0 0.3 0:00.00 bash 7724 root 20 0 17164 1324 1052 R 0 0.2 0:00.01 topI think you have answered for yourself "The only difference is one is the master and the other the slave in MySQL."
What is “cached” in the top command?
http://askubuntu.com/questions/198549/what-is-cached-in-the-top-command
@xDragonZ is correct.
@xDragonZ - does that really result in a 300 odd MB increase in RAM? If so i'm happy to have solved it - I just can't fathom why it would need so much more, considering the slave holds the same data?
My first time running a MySQL replicated DNS server - so i'm merely curious as to the difference in resources required.
Thanks,
Is this on OpenVZ?
@NickM - Yes both OpenVZ
Is there anything missing from your top output? I don't even see pdns on NS1.
@NickM - Hmmm somehow it PDNS stopped itself on NS1 - started it again. Still usage on NS2 is far less - must be MySQL then as, as soon as I stop it usage falls dramatically.
So it seems master VPS needs way more specs than the slave.
Would one be able to run master on say 512MB and Slave on 128MB or would that only work until the master falls over and the slave needs to pickup the slack?
OR
Would it be fine as the slave doesn't replicate back to master?
I can't really think of a reason that your master would be using that much more memory than your slave. I suppose if your ns2 isn't getting any traffic at all, your ns1 would be using a bit more memory.
I really wouldn't recommend anything less than 512MB for a MySQL server, regardless of whether it's a slave or a master. Sure, you can tweak mysql to the point where it's using very little memory, but at that point, you're just going to be hitting the disk more as tables no longer fit into memory, etc.
Actually... I just thought of something... Is it possible that one node is using 2.6.18 kernel and the other is using 2.6.32? OpenVZ counts memory differently depending on which kernel is being used, so you might want to check with your provider (uname -a will NOT tell you the actual kernel version being used on the node).
Aside: Is anyone even using .18 anymore?
Slightly off-topic: I know this is LowEndTalk, but I've never been one to worry about memory usage too much. And nowadays, considering that you can get 1GB for $5 a month from reliable providers, I care even less.
Linux memory is basically
used = used
free = free + buffers + cache
Think of it this way.
Ideally, you get NOTHING from ram being free (if anything your host if OpenVZ gets to oversell from you not using your ram). So its best it is put to use. Linux puts it to use caching the file system (disk = slow, ram = fast).
BTW if you hit the ">" key in top it will change the sort by column alowing you to see what is using the memory.
Master has table and query caches, slave does not - at least not with row repl.
Try
mysqlreportto get more details on how MySQL is using RAM.Maybe your queries runs on the master and consuming bufferbspace.
Linux also uses memory for inode caching, which is not accounted for in your "cached" memory usage.
You need to do a sync && drop caches to see actual, real memory usage.