Suggestion for VS1

jscieza

I love KH
Hi,

IMHO, VS1 has a very limited amount of RAM and disk space. 192MB doesn't appears to be a standard amount of RAM for common introductory packages in the VPS industry.

My suggestion is to make the VS1 to have 256MB or 384MB of RAM and 20GB of disk space. That way it will be more usable :)

Thank you,
Jonathan
 
I hear what you are saying but you have to consider that for 10 dollars more you get 768MB RAM. And you are not going to get a managed VPS with 384MB RAM from any other host on par with KnownHost for $25 a month.
 
Although I doubt it will happen but I do agree ... 192MB is a bit too low. You can't even run "cP DNS only" properly on it.

384MB (or even 256MB) sounds more like it.

No harm in suggestion .. right :)
 
Absolutely no harm in asking :)

But I do have to say that KH makes it known that some control panels should not be run on the VS1.

** Plesk, DirectAdmin or cPanel control panel is not recommended with 192 MB of RAM due to the amount of resources required for control panel itself and all services installed and managed by the control panel
 
I'm just curious. VS1 comes with 10GB of disk space. After installing the OS (without control panel), how much usable space remains to the user?
 
Depends on how you use it :)

remove "cp" - and 192 MB would be more than enough for "DNS only". Unless you have thousands of zones - and even in this case you could retire bind and use something less hungry. From my last experience with "bloated" bind (some 80000 zones) it didn't eat all the 2 GB memory available on box but took a considerable
time to restart.

Example:

small network, with some 8GB/day incoming & outgoing traffic combined, asymmetric link.

router does (in addition to routing :) :
1) packet filtering:

===
INFO:
Status: Enabled for 2 days 02:12:36 Debug: err

Interface Stats for re0 IPv4 IPv6
Bytes In 70667551916 0
Bytes Out 47492376126 64
Packets In
Passed 79578393 0
Blocked 325121 0
Packets Out
Passed 71936170 1
Blocked 17087 0

State Table Total Rate
current entries 108
searches 304282633 317.5/s
inserts 1205128 1.3/s
removals 1205020 1.3/s
Counters
match 1907314 2.0/s
bad-offset 0 0.0/s
fragment 148 0.0/s
short 78220 0.1/s
normalize 114 0.0/s
memory 0 0.0/s
bad-timestamp 172 0.0/s
congestion 0 0.0/s
ip-option 0 0.0/s
proto-cksum 1404 0.0/s
state-mismatch 4329 0.0/s
state-insert 626 0.0/s
state-limit 0 0.0/s
src-limit 27019 0.0/s
synproxy 0 0.0/s


2) packet scheduling (see asymmetric link above)
3) logging/stats through pfstat/pflogd
4) daemons running:

squid (as transparent proxy - all http/https traffic outside goes through squid)
exim
nsd
unbound
dhcpd
ntpd
httpd

+ regular stuff like logger, sshd, ftp-proxy, crond, inetd

100 MB of memory used, no swapping whatsoever:
root@gw:~# vmstat -w 1
procs memory page disk traps cpu
r b w avm fre flt re pi po fr sr wd0 int sys cs us sy id
1 0 0 101908 2645184 14 0 0 0 0 0 1 572 2287 26 1 4 94
0 0 0 101912 2645180 23 0 0 0 0 0 0 3247 126 25 0 2 98
0 0 0 101912 2645180 8 0 0 0 0 0 0 3065 37 16 0 1 99
0 0 0 101912 2645180 7 0 0 0 0 0 0 2733 33 16 0 1 99
0 0 0 101912 2645148 7 0 0 0 0 0 2 1982 52 22 0 0 100
^C
root@gw:~#

I always say that load average in vast majority of cases is not related to the
real load. The community think it does - well, here it is:
root@gw:~# w
1:28AM up 11 days, 2:21, 2 users, load averages: 0.19, 0.12, 0.09
USER TTY FROM LOGIN@ IDLE WHAT
barmaley p0 ##classified## Sat10PM 0 screen -r ###
moorzik p2 ##classified## Tue10PM 3:21 -authpf: ####@192.168.10.36
root@gw:~#

it's on intel atom, running on slowest clock speed

root@gw:~# apm
Battery state: absent, 0% remaining, unknown life estimate
A/C adapter state: not known
Performance adjustment mode: cool running (216 MHz)
root@gw:~#

So 192MB could be more than enough for a real wide range of usage scenarios which don't involve control panels or memory hogs like java and friends.
 
Top