[CentOS] iSCSI + kickstart

[Ugo Bellavance]

Hi,

(I'm using RHEL, but I posted on redhat.general and rhel5 mailing lists 
and didn't get an answer, so I'm taking a chance here as well)

Two of my servers have 2 nics (HP DL495c G5).  Using RHEL 5.4, I'd like 
to do a network install to boot from iSCSI, as the NIC supports it, 
following this document:  http://kbase.redhat.com/faq/docs/DOC-6875.

However, since my data network is on a network segment that is separated 
from the SAN (iSCSI) network segment, I'd need to install from pxeboot 
using the LAN VLAN, but configure another nic so that it has access ot 
the SAN VLAN.  From what I can see, it is not possible to assign a vlan 
in kickstart, and kickstart can only configure one NIC using the 
--network parameter.  However, I managed to configure another nic using 
a %pre%.  It also looks impossible to create a bonded interface in 
kickstart, which I'd like to be able to create.  Ideally, I'd like a 
bonded interface in failover mode, and some vlan'd interface on top of 
the bonded interface.  But I could manage that afterwards, I guess.

It looks like only my first physical NIC can boot from the SAN, so 
ideally, I'd need to have a eth0 to be in the SAN segment, and configure 
eth1 for the LAN, so that it can see my dhcp, pxeboot and tftp server.

Then I guess I must change the boot mode of my NIC so that it boots 
using iSCSI instead of PXE (after the install).  But then, will 
multipath be configured?  Since I have two controllers, the lun shows me 
two disks...

If anyone has had experience with this kind of installation, please let 
me know.

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Recent openvz kernels unstable?

[Ugo Bellavance]

On 2010-01-21 11:35, Chris wrote:
> On Thu, Jan 21, 2010 at 09:09:59AM -0500, Chris wrote:
>> Has anyone else running heavily loaded openvz boxes noticed strange behavior
>> with the latest kernels?  Ever since updating to the latest *.164* kernels, 
>> we
>> have mysterious machine lock ups.  Sometime there will be an oops message,
>> other times there will be none.  The oops messages aren't the same when we
>> manage to get one.  Sometimes a machine will just drop off network without
>> even a blip on the serial console.
>>
>> I still have to try to get a memory dump from one of the machines to see if I
>> can find anything there, but I'm curious if anyone else has run into the same
>> issues.
>
> Never mind.  I remembered after I sent this that ovz-kernel isn't a centos
> kernel, it just looks like one.  I also found
> https://bugzilla.redhat.com/show_bug.cgi?id=545411

Fixed in kernel-2.6.18-164.11.1.el5 
https://rhn.redhat.com/errata/RHSA-2010-0046.html

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Find reason for heavy load

[Ugo Bellavance]

On 2009-12-31 15:13, Noob Centos Admin wrote:
> Just an concluding update to anybody who might be interested :)
>
> My apologies for blaming spamassassin in the earlier email. It was
> taking so long because of the real problem.
>
> Apparently the odd exim processes that was related to the mail loop
> problem I nipped was still the culprit. I had overlooked the fact that
> by the time I caught onto the mail loop issue, there were actually
> hundreds if not thousands of bounced and rebounced messages in the
> queue already. Attempting to deliver these messages queued before I
> terminated the mail loop was what those exim processes were trying to
> do.
>
> This would had been ok if not for the other problem. The user
> apparently went on 2 week vacation since 15th and thought it was a
> good idea to enlarge his mailbox before doing so. So there was this
> 2.5GB mailbox choked full of both valid&  rebounced mails, plus the
> queue of more rebounced mails. So every time exim attempted to add the
> queued mails to the user's account, the quota system rejected it. The
> cpu load was probably due to this never ending ping pong match between
> exim and the quota.
>
> Yeah, I can't help but feel this must be such a noob mistake allowing
> that to develop without realizing it.
>
> Now that I've purged the queue of those bounced messages and other
> housekeeping for that user, server load has finally gone back to the
> expected sub 1.0 levels so I can finally go and enjoy my holiday :)
>
>
>
> On 1/1/10, Noob Centos Admin  wrote:
>> I initiated services shutdown as previously planned and once the
>> external services like exim, dovecot, httpd, crond (because it kept
>> restarting these services), the problem child stood out like a sore
>> thumb.
>>
>> There was two exim instances that didn't go away despite service exim
>> stop. Once I killed these two PID, the load average started dropping
>> rapidly. After a minute or so, the server went back to a happy 0.2~0.3
>> load and disk activity became almost negligible.
>>
>> I think these, orphaned? zombied?, exim instances were related to a
>> mail loop problem I discovered earlier today where one of my client on
>> holiday had a full mailbox and keep bouncing mails from a contact
>> whose site was suspended. Although I terminated that loop, it seemed
>> that exim had gotten those two instances stuck in limbo sucking up
>> processing power and hitting the disk somewhere unknown since they
>> weren't showing up in my exim logs.
>>
>> After observing a while, I brought the services back and once exim got
>> started, my load went back to 2.x ~ 3.x. Unfortunately while I was
>> typing this email, I realize it didn't stop there. I'm up to 4.x ~ 5.x
>> load level by now.
>>
>> So the application that is the cause of the load is definitely exim,
>> more specifically I think it's spam assassin because now that the mail
>> logs entries are slow, I can read the spamd details and mails are
>> taking between 3 to 8 seconds to be checked.
>>
>> Thanks again to everybody who had offer suggestions and advice and do
>> have a Happy New Year :)
>>
>>
>> On 1/1/10, Noob Centos Admin  wrote:
>>> Hi,
>>>
 I do not know about now but I had to unload the modules in question.
 Just clearing the rules was not enough to ensure that the netfilter
 connection tracking modules were not using any cpu at all.
>>>
>>> Thanks for pointing this out. Being a noob admin as my pseudonym
>>> states, I'd assumed stopping apf and restarting iptables was
>>> sufficient. I'll have to look up unloading module later.
>>>
 /me shrugs. When I was the mta admin at Outblaze Ltd. (messaging
 business now owned by IBM and called Lotus Live) spammers always ensured
 I got called. All they do is just press the big red button (aka start
 the script/system) and then go and play while I would have to deal with
 whatever was started.
>>>
>>> Based on the almost precise timing of around 9:30 to 5:30 India time,
>>> I'm inclined to think in my case it wasn't so much a spammer pressing
>>> a red button but a compromised machine in an office starting up when
>>> the user gets into office and knocks off on time at 5:30 :D
>>>
 I remember only one occasion when the spams were
 launched but neutralized very soon because they were pushing a website
 and I found a sample real early and so the anti spam system could just
 dump the spams and knock out accounts being used to send the crap.
>>>
>>> Could I ask how do I knock out the accounts sending the crap if they
>>> are not within my systems?
>>>
 First, try rmmod'ing the netfilter modules after you have cleared away
 the state related rules to make sure that you are only using static
 rules in netfilter...unless you have done that already..
>>>
>>> I think I'm only using static rules because after I restart iptables,
>>> I would then do a service iptables status to check my rules were in,
>>> and that list was very short c

Re: [CentOS] Find reason for heavy load

[Ugo Bellavance]

On 2009-12-29 23:44, Noob Centos Admin wrote:
> My Centos 5 server has seen the average load jumped through the roof
> recently despite having no major additional clients placed on it.
> Previously, I was looking at an average of less than 0.6 load, I had a
> monitoring script that sends an email warning me if the current load
> stayed above 0.6 for more than 2 minutes. This script used to trigger
> perhaps once an hour during peak periods. Even so, I seldom see numbers
> higher than 1.x
>
> On 4th Dec, somebody from an Indian IP range started hammering my SMTP
> service, attempting to use it as an open relay. Naturally that didn't
> work and only end up budging my typical 400KB daily log report into
> 2MB~4MB affairs.
>
> After observing a few days to determine the IP range, I started blocking
> the Indian subnet with apf. Initially I had problems with getting apf to
> wok properly but after a couple of days managed to get the block working
> and my daily log went back down to expected size when all those
> connection attempts disappear from exim's log.
>
> Now this is when my server load started to shoot through the roof with
> figures like 8.64 5.90 3.62 being reported by my monitoring script,
> triggering so often. I had to raise my threshold to 1.6 to keep my own
> script from spamming myself.
>
> I've tried changing several things on the server, since initially it
> seems like the high load may be due to I/O wait. So I turning off
> non-essential services like OpenNMS to see if that had any effect. I
> also turned off apf and inserted rules manually into iptables to reduce
> the number of iptable rules the system has to process.
>
> All that doesn't seem to help much, I'm still getting consistent server
> loads in the 2.x to 3.x range almost all the time.
>
> The problem is using top, none of my processes are showing abnormal
> CPU%, most are well under 5%, manually adding them up doesn't equate the
> 200% to 300% the load figures of 2.x and 3.x are indicating.
>
> Even top's own summary says CPU % is in the 20~30% range, what's
> worrying is the System% is also in the same range. I have no idea what
> is "system" doing since it appears that anything running inside the
> kernel is lumped under "system". Or why even totalling both % up, I
> would expect 50~60% to translate to the expected load of 0.5~0.6 yet
> system load stats is 5x what's expected.
>
> I've installed utilities like dstat to try to see if I can figure out
> which process is making the system calls that is clogging up the server
> but either I don't understand it or it's not the right tool.
>
> So I'll appreciate some advice on how/what should I do next to identify
> the cause. Thanks in advance!

Dstat could at least tell you if your problem is CPU or I/O.

Even better, run

vmstat 2 10

Look at the first two columns.  What column have higher numbers?  If r, 
you're CPU-bound.  If b, you're I/O bound.

If you're I/O bound, I suggest you use atop to determine which processes 
take disk time.

You can also use iostat -x 2 10.

I really suggest you read on vmstat and iostat, they will always be helpful.

Did you check if you have a defect disk or a rebuilding array?  That 
could be the cause.

Regards,

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] mkdir this "." directory

[Ugo Bellavance]

On 2009-12-28 18:49, adrian kok wrote:
> Hi
>
> I have this . folder under tmp

It is a system-generated link to the current directory.  Don't touch that.

Ugo


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] xen and hwclock

[Ugo Bellavance]

Mintairov Mihail a écrit :
> Hello All, I have a problem with hwclock on centos 5.3. in xen guest 
> environment.
> 
> # hwclock --debug
> hwclock from util-linux-2.13-pre7
> hwclock: Open of /dev/rtc failed, errno=19: No such device.
> No usable clock interface found.
> Cannot access the Hardware Clock via any known method.
> 
> So, I'm afraid that I can not change my time by ntp.
> 
> In CentOS mailing list I found a thread "hwclock and util-linux on 5.3", 
> but there is still no answer and discussion stopped. Maybe someone knows 
> the solution of this problem.

Maybe the clock is sync'd with the host?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Weird performance problem

[Ugo Bellavance]

JohnS a écrit :
> On Sun, 2009-04-19 at 09:02 -0700, Akemi Yagi wrote:
>> On Sun, Apr 19, 2009 at 8:34 AM, JohnS  wrote:
>>
>>> I don't reacall any one really saying if there was indeed a fix put into
>>> that PAE or any Specific Kernel for CentOS. Can the CentOS Kernel
>>> Builder Comment Please?
>>>
>>> Also See: +1
>>> http://wiki.centos.org/TipsAndTricks/VMWare_Server?highlight=(100hz)
>>>
>>> JohnStanley
>> As noted at the top of that wiki page, the contents need to be
>> updated.  When I added that note, I intended to do it asap but have
>> not had a chance to do so.  However, the link referenced in there (
>> http://kb.vmware.com/kb/1006427 ) is the best source for timekeeping
>> at this moment.  In short, CentOS no longer offers 100Hz kernels
>> because the divider=10 kernel option now works.
>>
>> Akemi
> 
> Thanks Akemi for the update on it. That should fix hin up,

I'm not using VMWare, I'm using OpenVZ...

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Weird performance problem

[Ugo Bellavance]

JohnS wrote:

> -
> That's a known problem with the Kernel and VM Kernel. You need the fixed
> kernel.

Do you have an URL of the bug or something?

I updated to the latest kernel.

I was running 2.6.18-92.1.18.el5.028stab060.2PAE.

Regards,

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Weird performance problem

[Ugo Bellavance]

Ugo Bellavance a écrit :
> Hi,
> 
> I'm running a CentOS 4.  server and I sometimes face a weird problem. 
> It is a weird performance problem, and here is how I discovered it.
> 
> This server runs OpenVZ virtual machines, and one of them is an asterisk 
> server for my personal use.  The first symptom of the problem is that 
> the voice quality became flaky.  So I logged on the server to see what 
> could be eating cpu cycles, when I ran top, it took almost one minute 
> before top actually showed.  Another hint is that when I run dstat (a 
> monitoring utility that is a mix of iostat and vmstat and other stats), 
> I often get a "missed xx ticks", where xx is a number.

Another hint is that pings are really slow.  Even pinging localhost is 
very long.  The first reply is fast, but the second takes ages to come.

It seems to be blocking here:

recvmsg(3, 0xbfbf84b0, MSG_DONTWAIT)= -1 EAGAIN (Resource 
temporarily unavailable)
gettimeofday({1239887784, 389347}, NULL) = 0
poll(

The rest comes as soon as there is another response:

[{fd=3, events=POLLIN|POLLERR}], 1, 999) = 0
gettimeofday({1239887903, 119727}, NULL) = 0
gettimeofday({1239887903, 119791}, NULL) = 0
sendmsg(3, {msg_name(16)={sa_family=AF_INET, sin_port=htons(0), 
sin_addr=inet_addr("127.0.0.1")}, 
msg_iov(1)=[{"\10\0\335\2018)\0\4\0370\347I\357\323\1\0\10\t\n\v\f\r\16\17\20\21\22\23\24\25\26\27"...,
 
64}], msg_controllen=0, msg_flags=0}, MSG_CONFIRM) = 64
recvmsg(3, {msg_name(16)={sa_family=AF_INET, sin_port=htons(0), 
sin_addr=inet_addr("127.0.0.1")}, 
msg_iov(1)=[{"e\0\0t\26\264\...@\1e\363\177\0\0\1\177\0\0\1\0\0\345\2018)\0\4\0370\347I"...,
 
192}], msg_controllen=20, {cmsg_len=20, cmsg_level=SOL_SOCKET, 
cmsg_type=0x1d /* SCM_??? */, ...}, msg_flags=0}, 0) = 84
write(1, "64 bytes from hn01.domain"..., 82) = 82
recvmsg(3, 0xbfbf84b0, MSG_DONTWAIT)= -1 EAGAIN (Resource 
temporarily unavailable)
gettimeofday({1239887903, 120785}, NULL) = 0
poll(

Then it blocks again...

This confuses Nagios that is running in a VM on this server.

Can the 'gettimeofday' be the problem?  'date' runs w/o delay.

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Feature list between 5.2 and 5.3

[Ugo Bellavance]

Drew Weaver a écrit :
> does anyone know if (and where) a feature list of the changes between 
> CentOS 5.2 and 5.3 can be found?
> 
>  
> 
> All of the documentation that I could find on centos.org is related to 
> 5.2 and 5.1

Look for the release notes, and, at worse, look the upstream release notes.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Weird performance problem

[Ugo Bellavance]

Hi,

I'm running a CentOS 4.  server and I sometimes face a weird problem. 
It is a weird performance problem, and here is how I discovered it.

This server runs OpenVZ virtual machines, and one of them is an asterisk 
server for my personal use.  The first symptom of the problem is that 
the voice quality became flaky.  So I logged on the server to see what 
could be eating cpu cycles, when I ran top, it took almost one minute 
before top actually showed.  Another hint is that when I run dstat (a 
monitoring utility that is a mix of iostat and vmstat and other stats), 
I often get a "missed xx ticks", where xx is a number.

Example (current) (sorry for the wrap):

total-cpu-usage -dsk/total- -net/total- ---paging-- ---system--
usr sys idl wai hiq siq| read  writ| recv  send|  in   out | int   csw
   3   2  93   2   0   0| 106k  273k|   0 0 | 0.2   0.4 |1039   389
   3   6  91   0   0   0|   0  6416k| 276k  275k|   0 0 |2160  6822 
  missed 55 ticks
   4  10  84   2   0   0|1200k 1992k|  82k   93k|   0 0 |1188  6275 
  missed 29 ticks
   1   0  99   0   0   0|   0  1312k|  65k   66k|   0 0 |1050  1114 
  missed 38 ticks
   2   1  96   0   0   0|   0  1168k|  57k   59k|   0 0 | 491   877 
  missed 13 ticks
   3   1  94   1   0   0|   0  6016k| 181k  176k|   0 0 |2169  5996 
  missed 50 ticks
   4   2  91   1   0   0|  28k 8744k| 216k  214k|   0 0 |2159  5438 
  missed 37 ticks
   1   1  98   0   0   0|   0  2632k|  93k   91k|   0 0 | 983  1381 
  missed 34 ticks
   1   1  98   1   0   0|   0  5624k| 113k  110k|   0 0 |1569  2643 
  missed 52 ticks
   1   1  98   1   0   0|   0  2432k|  29k   28k|   0 0 | 679   647 
  missed 12 ticks
   0   0 100   0   0   0|   0 0 |  60B  374B|   0 0 |  1315
   2   3  94   0   0   0|   0  1872k| 209k  210k|   0 0 |1375  3590 
  missed 30 ticks



The problem is currently occuring, but it doesn't seem to be affecting 
voice quality for now, so I have some time to try to find the cause. 
The only solution I've found up to now is to reboot... But hey, this 
isn't a Windows 98 machine :)!

I tried restarting the VZ system, which restarts all the VMs, but it 
didn't solve the problem.  I can't tell if the problem occurs on a stock 
centos kernel, because the server is running production (but 
non-critical) virtual machines, so it is always running the openVZ kernel.

So here is what I've done for now:

- Top shows a load of about 0.4

- vmstat 1 10 shows this:

procs ---memory-- ---swap-- -io --system-- 
cpu
  r  b   swpd   free   buff  cache   si   sobibo   incs us 
sy id wa
  0  0592 191092 381720 5379560053684 3  3 
2 93  2
  0  0592 190720 381720 53795600 0 0   3260  1 
1 98  0
  0  0592 191092 381720 53795600 0 0   4159  0 
0 100  0
  1  0592 191092 381728 53794800 0  2584  31196 10 
4 66 19
  0  0592 189968 381732 53794400 0  2080  222   174  2 
3 79 16
  0  1592 189968 381732 53794400 0  3244  17073 10 
4 73 12
  0  0592 190216 381732 53794400 0   136   76   113  1 
2 93  4
  0  0592 189844 381732 53794400 0 0   3369  1 
1 98  0
  0  0592 189844 381732 53794400 0 0   2432  0 
0 100  0
  0  0592 190340 381732 53794400 0 0   2842  0 
0 100  0

iostat -x 1 (excerpt)

Device:rrqm/s wrqm/s   r/s   w/s  rsec/s  wsec/srkB/swkB/s 
avgrq-sz avgqu-sz   await  svctm  %util
sda  0.00 171.00  0.00 124.000.00 2368.00 0.00  1184.00 
19.10 0.141.13   0.02   0.20
sdb  0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
sdc  0.00 171.00  0.00 124.000.00 2368.00 0.00  1184.00 
19.10 0.171.35   0.02   0.30
sdd  0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
md0  0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
md2  0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
md1  0.00   0.00  0.00 294.000.00 2352.00 0.00  1176.00 
 8.00 0.000.00   0.00   0.00
dm-0 0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
dm-1 0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
dm-2 0.00   0.00  0.00 294.000.00 2352.00 0.00  1176.00 
 8.00 0.301.01   0.02   0.50
dm-3 0.00   0.00  0.00 294.000.00 2352.00 0.00  1176.00 
 8.00 0.301.01   0.02   0.50
dm-4 0.00   0.00  0.00  0.000.000.00 0.00 0.00 
0.00 0.000.00   0.00   0.00
dm-5 0.00   0.00  0.00  0.000.000

Re: [CentOS] Can not get IP address

[Ugo Bellavance]

Anand Vaddarapu a écrit :
> Hi,
> 
> while installing centos 5.2 i enabled dhcp. centos machine did not get 
> IP address and can not get to internet.
> 
> ifconfig results
> Link encap:Local Loopback
> inet addr:127.0.0.1 Mask:255.255.255.0
> inet6 addr:  ::1/128 Scope:Host
> UP Loopback RUNNING  MTU:16436 Metric:1
> RX packets:1251 errors:0 dropped:0 overruns:0 frame:0
> TX packets:212517 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0  txqueuelen:0

What is your internet connection?  If it is cable modem, reset your 
modem and try again.

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] OT iSCSI best practices

[Ugo Bellavance]

Ugo Bellavance a écrit :
> Hi,
> 
>   I have an HP MSA 2000i as SAN (iSCSI) and since I'm starting with SANs, 
> I'm seeking for advice.  We are currently using HP Proliant 380 G5 + MSA 
> 70 (SAS, direct-attached storage) for our production servers.
> 
>   At first, I thought I'd be using the SAN only for non-critical 
> operations, like temporary additional storage for disk-based backups and 
> snapshots, so I planned on using my existing network equipement to 
> connect the SAN. Now I'd like to use it as storage for some blade 
> servers that I want to use for virtualisation, so reliability is more 
> important.
> 
>   I'd just like to evaluate the risk, to make sure I plan correctly.  The 
> SAN is a dual controller, dual-port, with a mix of RAID10 and RAID 50 
> arrays, so this is pretty solid.  However, if I'm using only one switch 
> to connect the initiators to this SAN, is it more risky than relying on 
> the non-redundant RAID controllers that are in my HP servers?
> 
>   If it is recommended that I use multipathing, I guess that means 2 
> switches, additionnal nics but also another interconnect in the blade 
> enclosure?
> 
> If you need more information, just let me know :).

I realized this might be too OT, so I posted here...

http://forums13.itrc.hp.com/service/forums/questionanswer.do?threadId=1330854

There are interresting answers.

Regards and sorry for the noise.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] OT iSCSI best practices

[Ugo Bellavance]

Hi,

I have an HP MSA 2000i as SAN (iSCSI) and since I'm starting with SANs, 
I'm seeking for advice.  We are currently using HP Proliant 380 G5 + MSA 
70 (SAS, direct-attached storage) for our production servers.

At first, I thought I'd be using the SAN only for non-critical 
operations, like temporary additional storage for disk-based backups and 
snapshots, so I planned on using my existing network equipement to 
connect the SAN. Now I'd like to use it as storage for some blade 
servers that I want to use for virtualisation, so reliability is more 
important.

I'd just like to evaluate the risk, to make sure I plan correctly.  The 
SAN is a dual controller, dual-port, with a mix of RAID10 and RAID 50 
arrays, so this is pretty solid.  However, if I'm using only one switch 
to connect the initiators to this SAN, is it more risky than relying on 
the non-redundant RAID controllers that are in my HP servers?

If it is recommended that I use multipathing, I guess that means 2 
switches, additionnal nics but also another interconnect in the blade 
enclosure?

If you need more information, just let me know :).

Thanks in advance,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] Backup methods for an Oracle DB

[Ugo Bellavance]

nate a écrit :
> Ugo Bellavance wrote:
>> Hi,
>>
>>  I've been testing different methods and I'd like to have some advice.
>> I want to perform a cold backup once a week on the Oracle DB, and put it
>> on tape.  I'm using EMC Networker for backup software, and I am not too
>> at ease with the fact of doing eveything with Networker, because if
>> there is a problem with the backup, the Oracle DB might not come up
>> after the backup run.
> 
> What version and edition of Oracle?

Oracle Database 10g Enterprise Edition Release 10.2.x.x.x - 64bit Production
With the Partitioning and Data Mining options

> Use RMAN, that's what it's there for. You can backup online, or
> offline, full or incremental.

Well, we only use one main oracle server... the DBA says it is not worth 
the additionnal overhead.  I'm no Oracle guru.

> At my last company we ran Oracle 10gR2 standard edition connected to
> a small fiber channel SAN. I wrote a script that put the tables on the
> primary server in hotbackup mode, then snapshotted the Oracle volumes,
> and mounted the snapshots onto a virtual machine that was running
> software iSCSI. From there a job kicked off and ran RMAN to backup
> the database.

Ok, but is that the equivalent of doing a cold backup?

> Prior to that we ran enterprise edition and was able to run RMAN
> directly from the physical standby server. With standard edition
> you can't do that.

Ok

> The migration from Oracle EE to Oracle SE probably paid for the
> SAN in itself let alone the massive increases in productivity
> gained by the flexibility of a centralized storage system(copying
> production data went from ~2 days to about 1 hour, copying data
> to reporting database went from ~8 hours to ~10 minutes).
> 
> You can also run RMAN against the primary system as well(any edition
> I believe), though I didn't want to do that as it'd impact
> performance.

We don't really care about the performance, as we are ok with up to 
about 2 hours of complete downtime per week.

Thanks,

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Backup methods for an Oracle DB

[Ugo Bellavance]

Hi,

I've been testing different methods and I'd like to have some advice. 
I want to perform a cold backup once a week on the Oracle DB, and put it 
on tape.  I'm using EMC Networker for backup software, and I am not too 
at ease with the fact of doing eveything with Networker, because if 
there is a problem with the backup, the Oracle DB might not come up 
after the backup run.

So I thought of using disk-based backup.  I've tried scp'ing the files 
directly to my backup server, but the operation is too long (120 min). 
I tried generating a tar.gz directly to my backup server via SSH.  A 
decent 40 minutes, using mgzip (multi-thread gzip), 70 Gigs.  120 
minutes for a tar.bz2, using pbzip2 (parralel bzip2) (54 Gigs).  A tar 
sent directly to my backup server is quite huge (318 Gigs).  It is then 
taken to tape on the regular nightly backup.

My concerns are:

- Time needed to perform backup (downtime).
- Time needed to do a recovery.

For the backup, sending a tar.gz directly to the backup server seems to 
be the best option.  However, since I want to minimize the time needed 
to perform a recovery, I'd like to have raw files on tape, not in a 
tarball and not compressed (the tape is compressing anyway).

Up to now, I've been quite disappointed by the speed at which my backup 
server can decompress and untar, and this server has quite good 
hardware.  When I use iostat -x, I find that the %util of the device is 
averaging 80%.

Here is the hardware involved:

Oracle Server:

HP Proliant 380DL + MSA70
16 GB ram
2 x Quad-core Xeons E5345 2.33 Ghz
9 RAID 10 volumes on 32 72G, 15K rpm SAS disks
1X Smart Array P400 w/512 MB BBU
1X Smart Array P800 w/512 MB BBU

Database Server

HP Proliant 360DL
2 x Quad-core Xeons E5345 2.33 Ghz
1 RAID 5 volume on 6 146 GB, 10K rpm SAS disks
1X Smart Array P400 w/512 MB BBU

Any help or suggestions welcome.

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Slightly OT: How to learn UNIX

[Ugo Bellavance]

Peter Arremann wrote:

On Thursday 06 March 2008, Ugo Bellavance wrote:

Oh, great,  any ideas about other Unix flavors? (AIX, HP-UX)


If I were you, I would forget about AIX at least at the beginning until you 
are solid with Solaris and HP-UX. Yes, it has good market share, but it is 
too different from everything else. 


Ok

Until recently I was a lead over a 8 person Unix admin team and our policy was 
to always hire people with little experience and then move them up as they 
learn stuff. 

I usually started them on Solaris 9. It is the closest to Linux (which most 
had experience with). Then we went on to Solaris 10 - SMF and so on are a big 
step forward but you will still find a ton of pre Sol10 out there, so if you 
don't have Solaris 9 or prior experience, you're not quite there. 


Ok, what about opensolaris?  Is



Then, the next step is HP-UX. You can get a C3xx0 or J6xx0 on ebay for little 
money these days. PA-RISC is dead, but once the OS is booted, there aren't 
that many differences between running on Itanium or PA-RISC. 

Finally, if they got that far, we would add some AIX. AIX is very different 
from what you would expect in a Unix flavor, ODM and all, but then again, 
Solaris 10 has moved quite a bit away from being a traditional Unix too. 


Peter.


Thanks for your great advice,
Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Slightly OT: How to learn UNIX

[Ugo Bellavance]

Steve Huff wrote:


On Mar 6, 2008, at 7:02 AM, Ugo Bellavance wrote:

I was wondering what would be the best way to learn AIX, Solaris, 
or HP-UX, for someone who knows Linux very well?  Books?  Courses? 
Self-teaching in a home lab?


in addition to the other suggestions, i recommend a copy of Evi Nemeth's 
"Unix System Administration Handbook" 
(http://www.admin.com/Pages/USAH.html).  one of the distinguishing 
features of this book is that for each topic it provides configuration 
examples for several different UNIX variants, highlighting the 
differences and similarities.


the current edition is the Third; you may also be interested in the 
Second edition, which covers some more proprietary UNIX variants.


Thanks, but I find it hard to spend so much on a book that is 8-year old...

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Slightly OT: How to learn UNIX

[Ugo Bellavance]

Christopher Chan wrote:



I recommend you installing OpenSolaris in a virtual or physical
machine and build a labor environment.


Yes.



Buy study guides for Solaris certified System, Network and Security
Administrator (for Solaris 10 boxes). OpenSolaris is released under an
OSI approved license and pretty innovative.


:-O

Solaris/OpenSolaris comes with GREAT documentation. The only problem is 
reading through them. There is a lot of it!


I suggest taking things one step at a time. Don't try to go through the 
documentation all at once. Just look it up when you need to do a step on 
Solaris be it setup a interface or a nfs share and then repeatedly do 
that. Say ten times on day one. Then 5 times two days later. Once more a 
week later. Hopefully you get to do the same procedure once in a while 
afterwards. If all else fails, just hit the documentation.


http://www.opensolaris.org/os/community/documentation/

Links to System Admin/Storage/Security/younameit on the left. They are 
all downloadable so that you do not have to read them on sun's website 
and they are available in pdf should you fancy printing them out.


This are also the opensolaris mailing lists.


Oh, great,  any ideas about other Unix flavors? (AIX, HP-UX)

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Slightly OT: How to learn UNIX

[Ugo Bellavance]

Hi,

	I was wondering what would be the best way to learn AIX, Solaris, or 
HP-UX, for someone who knows Linux very well?  Books?  Courses? 
Self-teaching in a home lab?


Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: slow SMTP-AUTH

[Ugo Bellavance]

Sorry for top posting, but I found the "problem".

My laptop is part of a Win2000 domain, so the solution was to disable 
kerberos in thunderbird.


I found it was kerberos by sniffing the packets.

Thanks,

Ugo

Ugo Bellavance wrote:

Ray Van Dolson wrote:

On Wed, Jul 25, 2007 at 07:35:58PM -0400, Ugo Bellavance wrote:

Ugo Bellavance wrote:

Brett Schroeder wrote:

Ugo Bellavance wrote:

Hi,

   I installed sendmail on a centos-4 openvz VM.  I set up SMTP-AUTH
and when I try to send mail using this server, there is a big 
delay in
which my e-mail client (Mozilla Thunderbird) just hangs.  The 
server, on
the other hand, is idle at the same time.  It does that no matter 
if I
use no encryption, TLS, or SSL.  When the client is 'hung' 
sendmail, on

the server side haven't logged anything yet for this connection.

Any idea?


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Huge mailq

[Ugo Bellavance]

Jason Pyeron wrote:
 


-Original Message-
From: [EMAIL PROTECTED] 
[mailto:[EMAIL PROTECTED] On Behalf Of nate

Sent: Monday, February 25, 2008 21:47
To: centos@centos.org
Subject: Re: [CentOS] Huge mailq

Jason Pyeron wrote:

Where should we start on preventing this type of problem?

[EMAIL PROTECTED] mqueue]# find | wc -l
185259

/etc/init.d/sendmail stop
chkconfig --level 2345 sendmail off
find /var/spool/mqueue -type f -exec rm -f {} \;



Funny, 


But we tried an even esier rm -rf / && reboot


Eheh, but the '&& reboot' part is rather useless as the reboot symlink 
should have disappeared once the 'rm -rf /' has done its magic ;).


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Internet Load Balancing and Failover

[Ugo Bellavance]

Shawn Everett wrote:

Hi All,

Does anyone have any experience connecting two or more DSL/Cable modems to 
a Linux box to provide load balancing and failover?


I've done some googling and found a few resources but very few solid 
experiences.


I'm trying to optimize my LAN->Internet traffic for a bunch of 
workstations.


I'd suggest using PfSense for that.  It is based on FreeBSD, but using 
OpenBSD's PF as packet filter.  Commercial grade features in this free 
firewall, definitely a lot easier to configure and maintain than a linux 
box for this kind of tasks.


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: domain name display issue in linux pc

[Ugo Bellavance]

[EMAIL PROTECTED] wrote:



Hi,


Thanks for your response on the kernel switching. I was away and could 
not reply immediately.



Right now, I am facing a different issue. I have to set up DNS server 
using BIND on Centos 4.3. When I type the hostname on Centos, it shows:


sipserver.vodcalocal.com

But the cli prompt has [EMAIL PROTECTED]  ~ meaning 
only the sipserver part of the hostname is displayed. why is this so? 
What is the actual hostname then? I see in the


The actual hostname is sipserver, the domain name is vodcalocal.com, so 
the FQDN is sipserver.vodalocal.com.


The shell is configured by default to show only the hostname part.  It 
is configurable (see Thomas' post).


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Backing up remote system

[Ugo Bellavance]

Scott Ehrlich wrote:
I have an Overland Arcvault tape library, a CentOS 5 box, a Windows XP 
system, and RAID box that supports NFS and CIFS.


The RAID box is remotely located and acts as central file storage.

I might normally use dump to perform backups, but as was learned here, 
and on dump's man page, dump doesn't support remote file systems such as 
NFS or CIFS.


So, for now, I've connected the library to a Windows XP system, running 
Service Pack 2.  I have a shell script that tar gz's the directories of 
choice to preserve, places them in /backup, and, via samba's config of 
making /backup available, have it mounted as a drive letter on the XP 
system.  I'm then using XP's built-in backup/restore program to store 
the contents of /backup to tape, and that is working fine for now.


Is there a reliable Linux/CentOS-based way to do this, too?   I know 
people keep saying use Amanda and/or bacula and be done with it, but in 
the event something goes wrong and the mysql, etc database gets corrupt, 
and people want their data back, it would be much easier to use a 
readily-available command, like cpio, tar, etc, to do the job.


For mysql, I strongly using MySQL's mysqldump or a wrapper to backup the 
DB, then put it to tape.


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Update problem

[Ugo Bellavance]

Hi,

	I was doing updates on a server and I think that the network connection 
got reset.  So I guess the yum update didn't complete and I now get this 
when I run 'yum update':


===
Setting up Update Process
Setting up repositories
Reading repository metadata in from local files
Resolving Dependencies
--> Populating transaction set with selected packages. Please wait.
---> Package system-config-printer-gui.i386 0:0.6.116.10-1.2 set to be 
updated

---> Package system-config-packages.noarch 0:1.2.23-2 set to be updated
---> Package cvs.i386 0:1.11.17-9.RHEL4 set to be updated
---> Package kernel-hugemem-devel.i686 0:2.6.9-67.0.4.EL set to be updated
---> Package kudzu.i386 0:1.1.95.23-1 set to be updated
---> Package glibc-kernheaders.i386 0:2.4-9.1.100.EL set to be updated
---> Package procps.i386 0:3.2.3-8.9 set to be updated
---> Package postgresql-contrib.i386 0:7.4.19-1.el4_6.1 set to be updated
---> Package gnome-keyring.i386 0:0.4.0-1.2.EL4 set to be updated
---> Package tcpdump.i386 14:3.8.2-12.el4_6.1 set to be updated
---> Package at.i386 0:3.1.8-80_EL4 set to be updated
---> Package system-config-date.noarch 0:1.7.15-0.RHEL4.3 set to be updated
---> Package cyrus-sasl.i386 0:2.1.19-14 set to be updated
---> Package postgresql-libs.i386 0:7.4.19-1.el4_6.1 set to be updated
---> Package foomatic.i386 0:3.0.2-3.1 set to be updated
---> Package kdelibs.i386 6:3.3.1-9.el4 set to be updated
---> Package nss_ldap.i386 0:226-20 set to be updated
---> Package cups.i386 1:1.1.22-0.rc1.9.20.2.el4_5.2 set to be updated
---> Package dbus-glib.i386 0:0.22-12.EL.9 set to be updated
---> Package nfs-utils.i386 0:1.0.6-84.EL4 set to be updated
---> Package gd-devel.i386 0:2.0.28-5.4E set to be updated
---> Package kernel.i686 0:2.6.9-67.0.4.EL set to be installed
---> Package dbus-python.i386 0:0.22-12.EL.9 set to be updated
---> Package postgresql-server.i386 0:7.4.19-1.el4_6.1 set to be updated
---> Package samba-common.i386 0:3.0.25b-1.el4_6.4 set to be updated
---> Package system-config-lvm.noarch 0:1.0.23-1.0 set to be updated
---> Package authconfig-gtk.i386 0:4.6.10-rhel4.3 set to be updated
---> Package xorg-x11-xauth.i386 0:6.8.2-1.EL.33.0.2 set to be updated
---> Package bind.i386 20:9.2.4-28.el4 set to be updated
---> Package pciutils.i386 0:2.1.99.test8-3.4 set to be updated
---> Package nscd.i386 0:2.3.4-2.39 set to be updated
---> Package gnupg.i386 0:1.2.6-9 set to be updated
---> Package openssh-clients.i386 0:3.9p1-8.RHEL4.24 set to be updated
---> Package gcc-java.i386 0:3.4.6-9 set to be updated
---> Package curl-devel.i386 0:7.12.1-11.el4 set to be updated
---> Package cyrus-sasl-md5.i386 0:2.1.19-14 set to be updated
---> Package gnome-vfs2.i386 0:2.8.2-8.6.EL4 set to be updated
---> Package mc.i386 1:4.6.1-0.8.5 set to be updated
---> Package squid.i386 7:2.5.STABLE14-1.4E.el4_6.1 set to be updated
---> Package authconfig.i386 0:4.6.10-rhel4.3 set to be updated
---> Package yum-utils.noarch 0:0.5-2.el4.centos set to be updated
---> Package bind-libs.i386 20:9.2.4-28.el4 set to be updated
---> Package cups-libs.i386 1:1.1.22-0.rc1.9.20.2.el4_5.2 set to be updated
---> Package crypto-utils.i386 0:2.1-4.2 set to be updated
---> Package libwnck.i386 0:2.8.1-7.el4 set to be updated
---> Package qt.i386 1:3.3.3-13.RHEL4 set to be updated
---> Package logwatch.noarch 0:5.2.2-2.EL4 set to be updated
---> Package NetworkManager.i386 0:0.3.1-4.el4 set to be updated
---> Package postgresql-pl.i386 0:7.4.19-1.el4_6.1 set to be updated
---> Package pam_krb5.i386 0:2.1.17-1 set to be updated
---> Package pam_ccreds.i386 0:3-3.rhel4.2 set to be updated
---> Package bluez-utils.i386 0:2.10-2.2 set to be updated
---> Package ntp.i386 0:4.2.0.a.20040617-6.el4 set to be updated
---> Package cyrus-sasl-plain.i386 0:2.1.19-14 set to be updated
---> Package dhcpv6_client.i386 0:0.10-17_EL4 set to be updated
---> Package openssl-devel.i586 0:0.9.7a-43.17.el4_6.1 set to be updated
---> Package libxml2-devel.i386 0:2.6.16-10.1 set to be updated
---> Package mdadm.i386 0:1.12.0-5 set to be updated
---> Package mkinitrd.i386 0:4.2.1.13-1 set to be updated
---> Package emacs.i386 0:21.3-19.EL.4 set to be updated
---> Package bind-utils.i386 20:9.2.4-28.el4 set to be updated
---> Package python-urlgrabber.noarch 0:2.9.8-2 set to be updated
---> Package xorg-x11-Mesa-libGL.i386 0:6.8.2-1.EL.33.0.2 set to be updated
---> Package system-config-samba.noarch 0:1.2.21-1.el4.1 set to be updated
---> Package initscripts.i386 0:7.93.31.EL-2.centos4 set to be updated
---> Package krb5-workstation.i386 0:1.3.4-54 set to be updated
---> Package xorg-x11-font-utils.i386 0:6.8.2-1.EL.33.0.2 set to be updated
---> Package libxml2-python.i386 0:2.6.16-10.1 set to be updated
---> Package glibc-devel.i386 0:2.3.4-2.39 set to be updated
---> Package postgresql-tcl.i386 0:7.4.19-1.el4_6.1 set to be updated
---> Package curl.i386 0:7.12.1-11.el4 set to be updated
---> Package libgcj.i386 0:3.4.6-9 set to be updated
--

[CentOS] Re: Log Monitoring Recomendation

[Ugo Bellavance]

Bill Campbell wrote:

On Tue, Jan 08, 2008, Ugo Bellavance wrote:

Joseph L. Casale wrote:
Given my experience in Linux is limited currently, what do you guys use 
to monitor logs such as ?messages? on your centos servers? I had a 
hardware failure that happened in between me manually looking (of 
course?). I would hope it might have a some features to email critical 
issues etc?

logwatch is a good start.

Get the latest version from www.logwatch.org.  Runs automatically daily 
and sends output to root.


Isn't logwatch standard in CentOS installations?


Yes, but an outdated version.


Swatch monitors one or more log files in real time, with options
to report events immediately, or after some number of repeations
in a specified time period (e.g. report immediately if a network
interface goes into permiscuous mode, but only report something
else if there are ``n'' occurrences within a minute).

I've attached the swatchrc configuration file from this machine
which has several examples.


Thanks, I tried it once, but got swamped with e-mails.  I'll give it 
another try.  Is it good with big log files?  I tried the check_log 
plugin for nagios, but it generated way too much I/O and timed out most 
of the time.


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Switching To Raid1

[Ugo Bellavance]

Matt wrote:

I have this ASUS M2NBP-VM motherboard http://tinyurl.com/3xby3h
running CentOS 4.4 as a web/email server.  It has a 500Gb SATA2 drive
with about 32Gb in use.

The motherboard supports hardware raid.  Is there a way to switch to
RAID1 without reinstalling or loosing any data?


This is probably what we call 'fake raid'.  Better use software than 
trying to get it to work with linux.  End result will be the same anyway.




Also, if I am running raid how do I know if there is a failure on one
of the drives anyway?  Is hardware RAID1 a good idea?

Matt



In software raid, mdadm sends an e-mail to root when a disk fails.  In 
(real) hardware raid, it depends of the make/model, but most can be 
configured to send emails when there is a problem.

RAID1 is always a good idea, especially for servers.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Log Monitoring Recomendation

[Ugo Bellavance]

Joseph L. Casale wrote:
Given my experience in Linux is limited currently, what do you guys use 
to monitor logs such as ‘messages’ on your centos servers? I had a 
hardware failure that happened in between me manually looking (of 
course…). I would hope it might have a some features to email critical 
issues etc…


logwatch is a good start.

Get the latest version from www.logwatch.org.  Runs automatically daily 
and sends output to root.


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: DRBD NFS load issues

[Ugo Bellavance]

Jed Reynolds wrote:

Ugo Bellavance wrote:

Jed Reynolds wrote:

Jed Reynolds wrote:

Ugo Bellavance wrote:



Can you send us the output of vmstat -n 5 5
when you're doing a backup?



This is with rsync at bwlimit=2500



This is doing the same transfer with SSH. The load still climbs...and 
then load drops. I think NFS is the issue.


I wonder if my NFS connection settings in client fstabs are unwise? I 
figured with beefy machine and fast networking, I could take 
advantage of large packetsizes. Bad packet sizes?



Are you backing up nfs to nfs?  From where to where are you doing 
backups?


The source data is on a ext3 partition, LVM volume, backed by a 15krpm 
raid 10 volume. Both rsyncs where conducted from the source host (the db 
server) to the backup server (which hosts nfs). In the nfs backup, I was 
rsyncing from the db filesystem to an NFS mount, and the ssh backup, I 
was rsyncing from the db filesystem to [EMAIL PROTECTED]:/backups.


Jed


Ok, so the DB is residing locally on the local server you are copying 
onto the nfs mount only during backup?


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: DRBD NFS load issues

[Ugo Bellavance]

Jed Reynolds wrote:

Jed Reynolds wrote:

Ugo Bellavance wrote:



Can you send us the output of vmstat -n 5 5
when you're doing a backup?



This is with rsync at bwlimit=2500



This is doing the same transfer with SSH. The load still climbs...and 
then load drops. I think NFS is the issue.


I wonder if my NFS connection settings in client fstabs are unwise? I 
figured with beefy machine and fast networking, I could take advantage 
of large packetsizes. Bad packet sizes?



Are you backing up nfs to nfs?  From where to where are you doing backups?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Remote desktop...

[Ugo Bellavance]

Tolun ARDAHANLI wrote:

Hi;

I want to connect from linux to Centos;)

what should i use at user side and server side?


SSH :).

More seriously... x terminal, vnc, but inside an SSH tunnel would be 
even better.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: DRBD NFS load issues

[Ugo Bellavance]

Jed Reynolds wrote:

My NFS setup is a heartbeat setup on two servers running Active/Passive
DRBD. The NFS servers themselves are 1x 2 core Opterons with 8G ram and
5TB space with 16 drives and a 3ware controller. They're connected to a
HP procurve switch with bonded ethernet. The sync-rates between the two
DRBD nodes seem to safely reach 200Mbps or better. The processors on the
active NFS servers run with a load of 0.2, so it seems mighty healthy.
Until I do a serious backup.

I have a few load balanced web nodes and two database nodes as NFS
clients. When I start backing up my database to a mounted NFS partition,
a plain rsync drives the NFS box through the roof and forces a failover.
I can do my backup using --bwlimit=1500, but then I'm not anywhere close
to a fast  backup, just 1.5MBps. My backups are probably 40G. (The
database has fast disks and between database copies I see run at up to
60MBps - close to 500Mbps). I obviously do not have a networking issue.

The processor loads up like this:
bwlimit   1500   load2.3
bwlimit   2500   load   3.5
bwlimit   4500   load   5.5+

The DRBD secondary seems to run at about 1/2 the load of the primary.

What I'm wondering is--why is this thing *so* load sensitive? Is it
DRBD? Is it NFS? I'm guessing that since I only have two cores in the
NFS boxes that a prolonged transfer makes NFS dominates 1 core and DRBD
dominate the next, and so I'm saturating my processor.


Is your CPU usage 100% all the time?


Can you send us the output of vmstat -n 5 5
when you're doing a backup?

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Matt wrote:

   I will probably have to design an e-mail (and other components)
infrastructure for a small ISP soon (WISP).

   I'm doing some research to determine which components would be best to
offer e-mail services to their client and allow the staff to manage
accounts easily.

   I usually use virtual machines a lot for isolation and easy backups and
migration (when a hardware node is underpowered, it is easy to migrate
one or more virtual machines to another hardware node easily).

   I have looked at iSCSI and drbd for high-availability of the storage:
http://www.pcpro.co.uk/realworld/82284/san-on-the-cheap/page1.html.

This looks like it should be doing a great job of high availability storage.

   For mail server, I guess I should look at an MTA and IMAP/POP server
that supports LDAP and/or MySQL for users.  Postfix should be a good
choice for MTA, as I know it (at least a little, but I know sendmail
better).  For IMAP/POP, I'm not sure...  Would dovecot be sufficient, or
should I try cyrus.  I'd rather use components that are available for
base or extras repository (or rpmforge).  I think that squirrelmail and
horde would do a good job for webmail.

   There shoudn't be any troubles having some redundancy for DNS, web
servers, mtas, but what about IMAP/POP? linux-HA?  MySQL replication
should be enough, I guess.  Or maybe linux-HA as well.  I wonder if I
should add GFS to the mix to have multiple IMAP/POP servers use the same
storage.  Or maybe IMAP proxies?

   Any insights welcome :).


www.directadmin.com

Been running it on CentOS for years.  Added Clamav and spamassassin to
it.  It utilizes Exim and dovecot along with standard bind and apache
stuff.  You pay monthly or yearly license fee.  Its pretty cheap
really.  You can also pay a one time fee for a given machine.



Thanks

Does it offer some kind of high availability features?  Does it provides 
an API (for account creation/management from another system)?


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Matt wrote:

It's been awhile since I looked into it, but I recommend outsourcing your
email.

Companies like fusemail (http://www.fusemail.com/solutions/resellers.html )
will give you accounts at $0.69/month/account for a 1GB account (last time I
checked anyway). They provide an API and a dashboard for managing it. They
also handle backups, I believe (read the fine print before you trust me). I
think there is spam filtering and anti-virus built in too.


Kept thinking that as well.  Only issue is that 0.69 x 2000 equals a
pretty good chunk of change for us.  One thing I would reccommend
though is putting it in a colocation facillity rather then local.
Also start with a beefy machine because its a real pain to upgrade
later.


I don't know if they have space already, but I agree on colocation.  For 
the upgrade, I'll probably virtualize it, so it is easy to upgrade the 
host when needed.


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Bill Campbell wrote:

On Thu, Jan 03, 2008, Ugo Bellavance wrote:

Hi,



I will probably have to design an e-mail (and other components)
infrastructure for a small ISP soon (WISP).


See my previous post on sizing mail servers.  The setup there is
in use at several of our regional ISP customers, and has been
very solid.  It's a design that has evolved since we started
building and selling systems for ISPs in 1994.


Ok, but I can't seem to find it.  Can you send me the link (see the 
'archived-at' header in the message).



I'm doing some research to determine which components would be best to
offer e-mail services to their client and allow the staff to manage
accounts easily.


There are various tools available to do this.  I have set up very
restricted webmin configurations so the support people at the ISP
could do the necessary things easily with minimal chance of major
screwups (after I've patched some things in webmmin that allowed
it to remove /home when somebody typed in a bad directory :-).


Makes sense, but hard to integrate with billing, for example. But could 
do.  I'm trying to think long-term, while avoiding spending too much time/$.



I usually use virtual machines a lot for isolation and easy backups and
migration (when a hardware node is underpowered, it is easy to migrate one
or more virtual machines to another hardware node easily).



I have looked at iSCSI and drbd for high-availability of the storage:
http://www.pcpro.co.uk/realworld/82284/san-on-the-cheap/page1.html.



This looks like it should be doing a great job of high availability
storage.



For mail server, I guess I should look at an MTA and IMAP/POP server that
supports LDAP and/or MySQL for users.  Postfix should be a good choice for
MTA, as I know it (at least a little, but I know sendmail better).  For
IMAP/POP, I'm not sure...  Would dovecot be sufficient, or should I try
cyrus.  I'd rather use components that are available for base or extras
repository (or rpmforge).  I think that squirrelmail and horde would do a
good job for webmail.


The systems we build have postfix/amavise/clamav, courier-imap,
and usually horde/imp for webmail.  I personally don't like Cyrus
as I prefer to use standard Maildir which allows easy clustering
for mail delivery and IMAP/POP access.


Ok, what do you mean by easy clustering?  Over GFS?  What make it hard 
to cluster in Cyrus?



There shoudn't be any troubles having some redundancy for DNS, web servers,
mtas, but what about IMAP/POP? linux-HA?  MySQL replication should be
enough, I guess.  Or maybe linux-HA as well.  I wonder if I should add GFS
to the mix to have multiple IMAP/POP servers use the same storage.  Or
maybe IMAP proxies?

Any insights welcome :).


I hope you're not charging your client for your learning curve.


Of course not.  Actually, I'm doing research even before having the 
requirements, to know what is possible as soon as possible, once I have 
the requirements, and to be able to discuss with the potential client 
without having to say "I'll have to research on this" every 5 seconds.


Thanks a lot for your help :).

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Ralph Angenendt wrote:

Gary Richardson wrote:

It's been awhile since I looked into it, but I recommend outsourcing your
email.


Erks. I wonder why *anyone* in his sane mind would do so (okay, here it
is smallish ISP but I - as a customer - trust my ISP to handle my mail
and would get another ISP as soon as I knew that it is outsourcing mail).


Thanks for the input.  Other comments below.


   For mail server, I guess I should look at an MTA and IMAP/POP
server
that supports LDAP and/or MySQL for users.  Postfix should be a good
choice for MTA, as I know it (at least a little, but I know sendmail
better).  


Then why not use sendmail? Once it is configured properly, maintaining
users is the same as with other MTAs.


I know that it supports ldap, but does it support MySQL or another DBMS 
for addresses lookup?



For IMAP/POP, I'm not sure...  Would dovecot be sufficient, or
should I try cyrus.  I'd rather use components that are available for
base or extras repository (or rpmforge).  I think that squirrelmail and
horde would do a good job for webmail.


I've heard that dovecot scales pretty good. If you want to be on the
safe side, cyrus probably scales way beyond what you need, but is also
harder to maintain.


Ok, but does it scales good like in one server can handle a lot or that 
it is easy to have multiple dovecot servers serve one domain in a 
transparent manner?



   There shoudn't be any troubles having some redundancy for DNS, web
servers, mtas, but what about IMAP/POP?


If you're already thinking about drbd - why not share the imap store
also?


You're right. But I'd need to have a second server ready to take the 
load if the first crashes.



And: There still is the Cyrus Murder for bigger setups, which
allows for flexibility within IMAP frontend and backend servers.


I wasn't aware of the existence of Cyrus Murder, I'll look it up.




  Or maybe IMAP proxies?


See 
which should answer most of your questions regarding HA within an imap
setup.


Wow, that is great, thanks!

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Christopher Chan wrote:

Fajar Priyanto wrote:

On Friday 04 January 2008 10:30:32 Ugo Bellavance wrote:
AFAIK, redundancy for mail server seldom uses linux-ha/any other 
failover

stuffs. It is most common to use 'backup MX' in DNS settings. So, when
the main server in unreachable, the sender mail server would try to the
secondary MX through DNS query.

That's the easy part, but where do you store the e-mail once you have
accepted it?  If the pop/IMAP server is down for a while, people won't
be able to retreive their e-mail...


The secondary MX will temporarily store the mails. And when the 
primary server is up again, it will get all the mail from the 
secondary. Yes. there will be a down time in terms of mail service for 
users.


Maybe others can recommend a better best practice for this.


Yes. No backup mx. You ought to have a cluster of mail servers to accept 
mails for your domain if you need HA. Otherwise, let incoming emails 
queue at their sending hosts as setting up a 'backup' mx that will only 
hold the email and then pass them onto the 'primary' is really pointless 
and only serves to 1) delay delivery of mail and 2) delay notification 
of mail delay to the sending party. It is no longer acceptable today to 
wait for a week before notifying the sender of non-delivery. The idea of 
a backup mx no longer fits today's communications.



I agree, and I don't want to have any client not being able to contact 
the pop/IMAP server for more than 15 minutes.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Gary Richardson wrote:
It's been awhile since I looked into it, but I recommend outsourcing 
your email.


Companies like fusemail 
(http://www.fusemail.com/solutions/resellers.html 
) will give you 
accounts at $0.69/month/account for a 1GB account (last time I checked 
anyway). They provide an API and a dashboard for managing it. They also 
handle backups, I believe (read the fine print before you trust me). I 
think there is spam filtering and anti-virus built in too.


Good idea, but that means no consulting fees for me ;).

I'll consider it, that is a fairly good idea.

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Mail server setup for small ISP

[Ugo Bellavance]

Fajar Priyanto wrote:

On Friday 04 January 2008 00:09:06 Ugo Bellavance wrote:

I think that squirrelmail and
horde would do a good job for webmail.

There shoudn't be any troubles having some redundancy for DNS, web
servers, mtas, but what about IMAP/POP? linux-HA?  MySQL replication
should be enough, I guess.  Or maybe linux-HA as well.  I wonder if I
should add GFS to the mix to have multiple IMAP/POP servers use the same
storage.  Or maybe IMAP proxies?

Any insights welcome :).


Hi Ugo,
Congrats on your plan opening the ISP.


Thanks, but I will only be doing some work for them, I'm not opening the 
ISP myself...


For webmail, a professional touch would be nice for your customer. Instead of 
using plain squirrelmail, try using http://www.nutsmail.com/ instead.


Good Idea

AFAIK, redundancy for mail server seldom uses linux-ha/any other failover 
stuffs. It is most common to use 'backup MX' in DNS settings. So, when the 
main server in unreachable, the sender mail server would try to the secondary 
MX through DNS query. 


That's the easy part, but where do you store the e-mail once you have 
accepted it?  If the pop/IMAP server is down for a while, people won't 
be able to retreive their e-mail...


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Mail server setup for small ISP

[Ugo Bellavance]

Hi,

	I will probably have to design an e-mail (and other components) 
infrastructure for a small ISP soon (WISP).


	I'm doing some research to determine which components would be best to 
offer e-mail services to their client and allow the staff to manage 
accounts easily.


	I usually use virtual machines a lot for isolation and easy backups and 
migration (when a hardware node is underpowered, it is easy to migrate 
one or more virtual machines to another hardware node easily).


	I have looked at iSCSI and drbd for high-availability of the storage: 
http://www.pcpro.co.uk/realworld/82284/san-on-the-cheap/page1.html.


This looks like it should be doing a great job of high availability storage.

	For mail server, I guess I should look at an MTA and IMAP/POP server 
that supports LDAP and/or MySQL for users.  Postfix should be a good 
choice for MTA, as I know it (at least a little, but I know sendmail 
better).  For IMAP/POP, I'm not sure...  Would dovecot be sufficient, or 
should I try cyrus.  I'd rather use components that are available for 
base or extras repository (or rpmforge).  I think that squirrelmail and 
horde would do a good job for webmail.


	There shoudn't be any troubles having some redundancy for DNS, web 
servers, mtas, but what about IMAP/POP? linux-HA?  MySQL replication 
should be enough, I guess.  Or maybe linux-HA as well.  I wonder if I 
should add GFS to the mix to have multiple IMAP/POP servers use the same 
storage.  Or maybe IMAP proxies?


Any insights welcome :).

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: [Semi-OT] Advice on large webmail setup

[Ugo Bellavance]

Michael Hill wrote:

www.atmaill.com
www.shupp.org
http://www.qmailtoaster.com

I use atmail with multi server mode


What procedures have you taken to make it high availability?

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: How to size an email server to handle 5 million emails per day

[Ugo Bellavance]

Erick Perez wrote:

I have no idea as to how to size an email server. I was approached by
a customer that wanted a single server with RAID 1 disks to handle
about 5 million emails a day.
In general terms, what parameters should I take into account to size
the hardware specs when the average email is about 10kb, the smalles
email is 2kb and the largest email is about 5meg (with attachment)

thanks,



I don't know if you have done it yet, but Fort System's new offering, 
BarricadeMX, could help you cut spam.


http://www.fsl.com/barricademx.html

It is closed-source, but FSL gives a lot to open-source communities, 
especially MailScanner's.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: OT: How many watts do I need?

[Ugo Bellavance]

Fajar Priyanto wrote:

Hello all,
Sorry for the OT thing.
I'm helping a friend setting up a new box. It's a Pentium 4 with 430Watt power 
supply. He wants to plugin his old harddrives and DVD into the new box. The 
total will be: 3 harddrives, 1 DVD drive, 1 CDRW drive. In the box there are 
also 2 big fans.


Is 430Watt enough?


What brand?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Firewall frustration

[Ugo Bellavance]

Robert Moskowitz wrote:

Firewall is up and running.

Used Shorewall with Webmin.

Les Bell wrote:

Robert Spangler <[EMAIL PROTECTED]> wrote:

  While IPTABLES might be CHEAP (price) it is a very good firewall.
Learn to set it up from the command line, it isn't that hard.
<<

Amen. I've been using CentOS for firewalls here for a long time now, with
hand-written rules. Besides, generic firewall configuration tools don't -
can't - know about many of the more advanced modules and features of
iptables.
I spent much of the past 24 hours trying to find out how to set up 
iptables for firewall routing WITHOUT NATing. Could not find anything.


So I decided to try out shorewall, which has a front end in Webmin. The 
'nice' thing about this was as I built a portion of Shorewall (say the 
zones), I could sue the Webmin edit the conf file directly to see the 
'raw' config file and looky there, a URL for a help page!


Taking it slow, I got Shorewall up in about 1 hour.

But I have questions for the Shorewall people. They talk about iptables, 
then netfilter. The site says that Shorewall is not a deamon. Well I see 
a Shorewall service running. Can't see that is using any cpu cycles or 
how much memory. The iptables have the same content they had when I used 
the upstream's tool at Centos install time to set up basic 'firewall' 
features. So what gives


There is also an iptables 'service', that doesn't mean there is a 
deamon.  It is a simple way to start the firewall at boot time.


Have you checked m0n0wall/pfsense livecd?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Firewall frustration

[Ugo Bellavance]

Mark Weaver wrote:

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, 1 Jan 2008 08:57:22 -0500
Robert Moskowitz <[EMAIL PROTECTED]> wrote:

Have you ever thought about how rare floppy drives are now?  At best
you go with a bootable usb, if your notebook supports bootable USB.
My Libretto does have a bootable floppy, but that is something extra
to carry.  It will not boot from anything else (besides its HD).  My
nc4010 (this notebook) will boot from usb.  My corp notebook (nc2400)
is locked down; and I don't see any value at getting corp IT bent out
of shape.


why would you even think about using a Notebook computer as a firewall?
I was assuming you were going to delegate this task to an older machine
with sufficient resources to handle the task and not give the task to a
notebook computer.


I guess he wants it to be portable.

He seems to be knowing his requirements a lot better than we do.  It 
looks like he wants an easy firewall that would boot for HD only, cost 
nothing, and runs with usb ethernet devices.


I really think he should carry an embedded firewall (like a soekris or a 
wrap) with pfsense on it.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: unziping

[Ugo Bellavance]

Christopher E wrote:

Hello All,

I am wondering what methoreds are allowed for unzipping a zip file
under CentOS 5?

How would you do this also?  I want to exact a file eo1.zip to
/home/dev/www/processing/eo1.lst <-- is the file name which comes from
the zip file.


the 'unzip' command should do it.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Help Needed with setup!

[Ugo Bellavance]

Christopher E wrote:

Hello Frank,

Yes it is "truly worth cause",  it is a 501c3 under the IRS of USA,
the project is to build a massive database that could be search to
find information about disabilities, like a person that is blind could
find funding for obtaining a computer that talks,  so the person on
the site would type in funding blind computer speech and it would find
the org that would help pay for such items, or agencies that help with
such thing,  person needing access-able housing could type wheelchair
housing and it would find place that help with doing such and even
bring up housing that is access-able also that have been added to the
database.


The project seems to be interesting, but I doubt anyone would volunteer 
for taking care of a server on a long-term basis.  Why just not get a 
PHP/MySQL shared hosting packages?  You an get one for less than 
10$/month. http://iweb8.com/web-hosting/ for example.  This is in a way 
more reliable than a single server, because they have a cluster of web 
servers and a cluster of MySQL servers, so they are a lot less likely to 
have downtime.  And this is managed, updated, backed up and so on.  The 
smallest package http://iweb8.com/web-hosting/micro/ should last for 
some time, and may fit in a permanent manner.  If you ever get more than 
5GB of data, the second package gives you 10GB for 12.77/month.


You can get even cheaper packages: 
http://www.godaddy.com/gdshop/hosting/shared.asp?isc=cjcophst20


The deluxe plan gives you 100GB space for 5.59/month!

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Torrent: reminder to use it folks!

[Ugo Bellavance]

[EMAIL PROTECTED] wrote:
Why use torrents? With torrents I get around 25Kb/sec. 


With places such as utah.edu [I am in North America] I got
320Kb/sec steady. It took me 3hr and a bit to download the 5.1
dvd. As far as I understand it, Utah and the other mirrors donated
the bandwidth to the community.



Even if they donate to the community, they still pay for it.  The less 
we take, happier they are to donate for many years.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: After 4.6 update: CUPS (via Samba): "'/printers/printers' no good!"

[Ugo Bellavance]

Bart Schaefer wrote:

On Dec 17, 2007 6:59 AM, Steve Snyder <[EMAIL PROTECTED]> wrote:

Any advice as to how to fix whatever was broken in the course of doing the
upgrade?


After any upgrade it's always good to run

updatedb
locate .rpmsave
locate .rpmnew

and then compare those files to the ones they might have replaced.


in fact only .rpmnew should have been generated.
Don't forget to merge your changes into the .rpmnew before replacing 
your current config.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: CentOS 5 and removing sendmail

[Ugo Bellavance]

Christian Volker wrote:

Yohoo!


# rpm -e sendmail
error: Failed dependencies:

rpm -e sendmail --nodeps

No --- Don't --- Install another MTA before you try to remove sendmail.


Where's the difference?



Using --nodeps can break stuff.

Instead to:

yum install postfix

then,

yum remove sendmail (if necessary)

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Good evening to all

[Ugo Bellavance]

Silambarasan.R wrote:


Hi..

Could u help me to re-install the centos 4 without affecting existing
data...? is it possible? if so pl let me know...

Regards
Silambu.R


Depends on your partition scheme.  If you used only one big / partition, 
probably not, but if you used a separate /home partition, you could save 
this one.


Why do you need to re-install?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: How to delete mails in the mailq in ONE DAY -- POSTFIX

[Ugo Bellavance]

Indunil Jayasooriya wrote:


I am runnig posfix on Centos 4.4 as a Mailgateway. It only accepts mails 
for domains and then forwards mails to Lotus domino Server.


Ok

All clients sends outgoing mails to that Lotus domino Server. Then , 
That Lotus Domino Server sends mails to Postfix mailgateway. This 
postfix mailgateway sends mails to all the destinations. But, This 
Postfix mailgateway has about 150 messages in the mailq. Some are 
unwanted mails. I think they keep trying until 5 days old. I want 
decrease it to ONE DAY.


These messages are not harmfull, and you should keep your setting to 5 
days.  You'll thank you when your domino server stops accepting messages 
for some reasons for 25 hours or more... it can happen...


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: RAID, LVM, extra disks...

[Ugo Bellavance]

Ross S. W. Walker wrote:

Ugo Bellavance wrote:

Ross S. W. Walker wrote:

What are you trying to accomplish storage wise?

Is this for commerical or personal use?

Commercial, but non-critical use.

If for personal use, then it isn't as critical how it is 
setup, but if 
this is for commercial use then you need to target your storage to 
your application.
Nothing really very IO demanding.  Running an OpenVZ server 
with many Virtual machines on it, but load is very low.


If you want this to be reconfigured on the fly without ever 
rebooting 
then you may find your options limited on which RAID levels you can 
choose.


Typically I keep the system disks in a RAID1 and the data disks on 
separate RAID arrays setup depending on the application.


Scratch or temp files -> RAID0
File serving -> RAID5 or RAID6 (depending on disk size # of disks) 
Databases, large email, many VMS -> RAID10


Let us know what you want the storage for and we can suggest a 
configuration.


Top of my head though, I would use the 18GB for the OS and 
keep the 4 

36GB for application data either as a RAID10 or RAID5.

That would make sense.  Use RAID1 18GB for /, /boot and /var and use a
RAID4 with 4 36GB HDD for /vz (OpenVZ's virtual machines are 
located there).


Makes sense?


Makes sense to me, I have found in my environment that VMs generate a
lot of random io, so a RAID10 may be better suited, though it means
72GB of useable space instead of 108GB.


I understand, but my IO is not very significant...  I think I'll use the 
RAID5 for /vz/



Also by using growing or sparse files for the vz images, a volume can get
fragmented pretty quickly. To minimzie that from happening, think about
creating LVs with separate small file systems to hold each vz image. If
the LVs start running out of space, you can grow them and the file
system as needed which will reduce the fragmentation tremendously.
You will still end up with LV extents fragmented, but since they are
larger it isn't as serious a performance issue.


OpenVZ doesn't work with vz image.  It works as a regular filesystem.

Thanks a lot for your advice.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: RAID, LVM, extra disks...

[Ugo Bellavance]

Ross S. W. Walker wrote:


What are you trying to accomplish storage wise?

Is this for commerical or personal use?


Commercial, but non-critical use.


If for personal use, then it isn't as critical how it is setup,
but if this is for commercial use then you need to target your
storage to your application.


Nothing really very IO demanding.  Running an OpenVZ server with many 
Virtual machines on it, but load is very low.




If you want this to be reconfigured on the fly without ever
rebooting then you may find your options limited on which
RAID levels you can choose.

Typically I keep the system disks in a RAID1 and the data disks
on separate RAID arrays setup depending on the application.

Scratch or temp files -> RAID0
File serving -> RAID5 or RAID6 (depending on disk size # of disks)
Databases, large email, many VMS -> RAID10

Let us know what you want the storage for and we can suggest a
configuration.

Top of my head though, I would use the 18GB for the OS and keep
the 4 36GB for application data either as a RAID10 or RAID5.


That would make sense.  Use RAID1 18GB for /, /boot and /var and use a 
RAID4 with 4 36GB HDD for /vz (OpenVZ's virtual machines are located there).


Makes sense?

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: RAID, LVM, extra disks...

[Ugo Bellavance]

Alain Spineux wrote:

On Nov 29, 2007 4:21 PM, Ugo Bellavance <[EMAIL PROTECTED]> wrote:

Alain Spineux wrote:

On Nov 29, 2007 6:59 AM, Ugo Bellavance <[EMAIL PROTECTED]> wrote:

Hi,

This is my current config:

/dev/md0 -> 200 MB -> sda1 + sdd1 -> /boot
/dev/md1 ->  36 GB -> sda2 + sdd2 -> form VolGroup00 with md2
/dev/md2 ->  18 GB -> sdb1 + sde1 -> form VolGroup00 with md1

sda,sdd -> 36 GB 10k SCSI HDDs
sdb,sde -> 18 GB 10k SCSI HDDs

I have added 2 36 GB 10K SCSI drives in it, they are detected as sdc and
sdf.

What should I do if I want to optimize disk space?

The simplest solution would be to create /dev/md3 out of sdc1 and sdf1,
and add it to the VG, and increase the size of my /vz logical volume.

However, if I could convert that to a RAID5 (it could be possible to
re-install, but I would rather not), I could have 6 drives in RAID5, so
I'd have 5x36 GB (180) of space available total, instead of 3*36 (108).

180, you mean 2 X 5x18

Oh, I just realized I have 2X18 and 4X36.  I have 2 other 36 GB HDD
here.  Maybe I could have a 6x36 RAID5 this way.  Does it matter if I


drop the 2x18, to drop the 2x18 complexity,
or keep them in mirroring, and maybe on another VG,
to backup/store more critical data


Ok, but the server can only have 6 HDD.

So, first step, how would I replace the current 2X18 by the 2x36 that I 
put in the server yesterday?


(BTW thanks a lot for your help, it is very interesting)

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: RAID, LVM, extra disks...

[Ugo Bellavance]

Alain Spineux wrote:

On Nov 29, 2007 6:59 AM, Ugo Bellavance <[EMAIL PROTECTED]> wrote:

Hi,

This is my current config:

/dev/md0 -> 200 MB -> sda1 + sdd1 -> /boot
/dev/md1 ->  36 GB -> sda2 + sdd2 -> form VolGroup00 with md2
/dev/md2 ->  18 GB -> sdb1 + sde1 -> form VolGroup00 with md1

sda,sdd -> 36 GB 10k SCSI HDDs
sdb,sde -> 18 GB 10k SCSI HDDs

I have added 2 36 GB 10K SCSI drives in it, they are detected as sdc and
sdf.

What should I do if I want to optimize disk space?

The simplest solution would be to create /dev/md3 out of sdc1 and sdf1,
and add it to the VG, and increase the size of my /vz logical volume.

However, if I could convert that to a RAID5 (it could be possible to
re-install, but I would rather not), I could have 6 drives in RAID5, so
I'd have 5x36 GB (180) of space available total, instead of 3*36 (108).


180, you mean 2 X 5x18


Oh, I just realized I have 2X18 and 4X36.  I have 2 other 36 GB HDD 
here.  Maybe I could have a 6x36 RAID5 this way.  Does it matter if I 
have 4 HDD that are 10K and 2 7200 rpm?


What about raid 6?  I don't think I need fault tolerance for 2 HDD 
failures...




Yes and without rebooting :-)

- break the 36GB mirror (using mdadm, faile /dev/sdd2, and then remove it),
- break the 18GB mirror (using mdadm, faile /dev/sde1, and then remove it),
- create sd[cf][123] of 200Mb, 18G, 18G (the 200Mb is useless but to
keep the same partitioning schema)
- create a _degraded_ raid5 with sd[cfd]2 sde1 named /dev/mdX
- vgextend your VolGroup00 to use this new partition.
# pvcreate /dev/mdX
# vgextend VolGroup00 /dev/mdX
- then move all PE  on md1  to mdX
# pvmove /dev/md1/dev/mdX
- then remove md1 from the VG
# vgreduce VolGroup00  /dev/md1
- now you dont need md1 anymore, stop it (sorry I'm less  skilled with
mdadm command, without manual page )
- now add /dev/sda2 to your _degraded_ raid 5


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: lvresize --resizefs

[Ugo Bellavance]

Matt Hyclak wrote:

On Wed, Nov 28, 2007 at 07:37:46PM -0500, Ugo Bellavance enlightened us:
	There is a difference between the help of lvresize and its man page. 
In the manpage, there is nothing about the -r or --resizefs function. 
Centos4.


[EMAIL PROTECTED] ~]# lvresize
  Please specify either size or extents (not both)
  lvresize: Resize a logical volume

lvresize
[-A|--autobackup y|n]
[--alloc AllocationPolicy]
[-d|--debug]
[-h|--help]
[-i|--stripes Stripes [-I|--stripesize StripeSize]]
{-l|--extents [+|-]LogicalExtentsNumber[%{VG|LV|FREE}] |
 -L|--size [+|-]LogicalVolumeSize[kKmMgGtTpPeE]}
[-n|--nofsck]
[-r|--resizefs]
[-t|--test]
[--type VolumeType]
[-v|--verbose]
[--version]
LogicalVolume[Path] [ PhysicalVolumePath... ]



Filing a bug in the PNAELV's bugzilla would be much appreciated, I'm sure.
We can't fix it here.


Done, bug 404581.

Thanks!

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

Christopher Chan wrote:

Ugo Bellavance wrote:

[EMAIL PROTECTED] wrote:

Christopher Chan pisze:


grub cannot find its second stage. Are you booting from a 
mirrored partition?


Yes


What could be a solution?  And what could have happen upon the 
reboot?


That is weird.  I just re-installed centos5 and it is now booting 
properly.  What could I do to avoid this situation in the future?




IIRC, RHEL4 does not properly handle installation of grub on 
mirrored partitions and therefore Centos4 suffers from the same 
problem.


RHEL5 does it properly now as you can see. This has been a long 
outstanding problem of anaconda.
Yeap, this is true. After installing centos4 on RAID1 disk (software 
raid) i always do:


grub
grub>device (hd0) /dev/hdc
grub>root (hd0,0)   grub>setup (hd0)

where /dev/hdc is second RAID DISK (it could be whatever: /dev/sdb1 etc)


Ok, on one system I had /boot as /dev/md0, and md0 is composed of 
/dev/hda1 and /dev/hdb1.


I have done:

grub> device (hd0) /dev/hdb

grub> root (hd0,0)
 Filesystem type is ext2fs, partition type 0xfd

grub> setup (hd0)
 Checking if "/boot/grub/stage1" exists... no
 Checking if "/grub/stage1" exists... yes
 Checking if "/grub/stage2" exists... yes
 Checking if "/grub/e2fs_stage1_5" exists... yes
 Running "embed /grub/e2fs_stage1_5 (hd0)"...  16 sectors are embedded.
succeeded
 Running "install /grub/stage1 (hd0) (hd0)1+16 p (hd0,0)/grub/stage2 
/grub/grub

.conf"... succeeded
Done.


Am I in the right way?


Looks good. The important part is that it references the drive when 
doing the installation.


What do you mean?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

Christopher Chan wrote:

Ugo Bellavance wrote:

[EMAIL PROTECTED] wrote:

Christopher Chan pisze:


grub cannot find its second stage. Are you booting from a 
mirrored partition?


Yes


What could be a solution?  And what could have happen upon the 
reboot?


That is weird.  I just re-installed centos5 and it is now booting 
properly.  What could I do to avoid this situation in the future?




IIRC, RHEL4 does not properly handle installation of grub on 
mirrored partitions and therefore Centos4 suffers from the same 
problem.


RHEL5 does it properly now as you can see. This has been a long 
outstanding problem of anaconda.
Yeap, this is true. After installing centos4 on RAID1 disk (software 
raid) i always do:


grub
grub>device (hd0) /dev/hdc
grub>root (hd0,0)   grub>setup (hd0)

where /dev/hdc is second RAID DISK (it could be whatever: /dev/sdb1 etc)

So system is booting form first or second riad1 disk


Ok, so to sum up from what I understand of my problem:

Installation of CentOS4 -> Installs grub only on one HDD partition out 
of 2, in the mirror.


Well, I believe you chose md0 or something to install grub on right?



One disk fails, the one that has grub


Not necessarily. All it takes is a change in the sequence of disk 
assignment.




System won't boot because it can't find grub on the other drive.


Correct. It needs to be instructed to look on its own drive.



If I had centos5 there in the first place, the setup would have taken 
care of installing grub on the 2 mirrored raid partitions.


Am I right?


Yes.



Is there a way to know where grub is installed?  I have a few servers 
running in RAID 1 software for /boot, I gotta fix this.  If I can't 
tell whether it is installed or not, is it dangerous to re-install it 
using the command above?




So long as you have all the necessary grub files, there is not much 
danger. Even if you are missing a config file on this side of the mirror 
(impossible...) so long as you load grub stage 2, you will have the 
power you need to continue if you have access (eg: via serial)


I was talking about all my others centos4 machines..

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

[EMAIL PROTECTED] wrote:

Christopher Chan pisze:


grub cannot find its second stage. Are you booting from a mirrored 
partition?


Yes


What could be a solution?  And what could have happen upon the reboot?


That is weird.  I just re-installed centos5 and it is now booting 
properly.  What could I do to avoid this situation in the future?




IIRC, RHEL4 does not properly handle installation of grub on mirrored 
partitions and therefore Centos4 suffers from the same problem.


RHEL5 does it properly now as you can see. This has been a long 
outstanding problem of anaconda.
Yeap, this is true. After installing centos4 on RAID1 disk (software 
raid) i always do:


grub
grub>device (hd0) /dev/hdc
grub>root (hd0,0)   grub>setup (hd0)

where /dev/hdc is second RAID DISK (it could be whatever: /dev/sdb1 etc)


Ok, on one system I had /boot as /dev/md0, and md0 is composed of 
/dev/hda1 and /dev/hdb1.


I have done:

grub> device (hd0) /dev/hdb

grub> root (hd0,0)
 Filesystem type is ext2fs, partition type 0xfd

grub> setup (hd0)
 Checking if "/boot/grub/stage1" exists... no
 Checking if "/grub/stage1" exists... yes
 Checking if "/grub/stage2" exists... yes
 Checking if "/grub/e2fs_stage1_5" exists... yes
 Running "embed /grub/e2fs_stage1_5 (hd0)"...  16 sectors are embedded.
succeeded
 Running "install /grub/stage1 (hd0) (hd0)1+16 p (hd0,0)/grub/stage2 
/grub/grub

.conf"... succeeded
Done.


Am I in the right way?

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

[EMAIL PROTECTED] wrote:

Christopher Chan pisze:


grub cannot find its second stage. Are you booting from a mirrored 
partition?


Yes


What could be a solution?  And what could have happen upon the reboot?


That is weird.  I just re-installed centos5 and it is now booting 
properly.  What could I do to avoid this situation in the future?




IIRC, RHEL4 does not properly handle installation of grub on mirrored 
partitions and therefore Centos4 suffers from the same problem.


RHEL5 does it properly now as you can see. This has been a long 
outstanding problem of anaconda.
Yeap, this is true. After installing centos4 on RAID1 disk (software 
raid) i always do:


grub
grub>device (hd0) /dev/hdc
grub>root (hd0,0)   grub>setup (hd0)

where /dev/hdc is second RAID DISK (it could be whatever: /dev/sdb1 etc)

So system is booting form first or second riad1 disk


Ok, so to sum up from what I understand of my problem:

Installation of CentOS4 -> Installs grub only on one HDD partition out 
of 2, in the mirror.


One disk fails, the one that has grub

System won't boot because it can't find grub on the other drive.

If I had centos5 there in the first place, the setup would have taken 
care of installing grub on the 2 mirrored raid partitions.


Am I right?

Is there a way to know where grub is installed?  I have a few servers 
running in RAID 1 software for /boot, I gotta fix this.  If I can't tell 
whether it is installed or not, is it dangerous to re-install it using 
the command above?


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

Ugo Bellavance wrote:

Ugo Bellavance wrote:

Christopher Chan wrote:

Ugo Bellavance wrote:

Hi,

I have a Dell poweredge 400SC here with an LSI scsi card.

I installed Centos4 a while ago and put it in a datacenter.  I 
rebooted many weeks after, and the machine didn't come back up.  So 
I went to the datacenter tonight to find out that the server was 
stuck at "GRUB". Nothing more.
So I decided to re-install Centos4.  The partitions were 
already there, so I just made sure partitions were formatted and the 
install went fine.  When I rebooted, "GRUB".


I'm at this point and I don't know exactly what I could do to 
diagnose furthere.


Any ideas?


grub cannot find its second stage. Are you booting from a mirrored 
partition?


Yes


What could be a solution?  And what could have happen upon the reboot?


That is weird.  I just re-installed centos5 and it is now booting 
properly.  What could I do to avoid this situation in the future?


Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

Ugo Bellavance wrote:

Christopher Chan wrote:

Ugo Bellavance wrote:

Hi,

I have a Dell poweredge 400SC here with an LSI scsi card.

I installed Centos4 a while ago and put it in a datacenter.  I 
rebooted many weeks after, and the machine didn't come back up.  So I 
went to the datacenter tonight to find out that the server was stuck 
at "GRUB". Nothing more.
So I decided to re-install Centos4.  The partitions were 
already there, so I just made sure partitions were formatted and the 
install went fine.  When I rebooted, "GRUB".


I'm at this point and I don't know exactly what I could do to 
diagnose furthere.


Any ideas?


grub cannot find its second stage. Are you booting from a mirrored 
partition?


Yes


What could be a solution?  And what could have happen upon the reboot?

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Stuck at GRUB

[Ugo Bellavance]

Christopher Chan wrote:

Ugo Bellavance wrote:

Hi,

I have a Dell poweredge 400SC here with an LSI scsi card.

I installed Centos4 a while ago and put it in a datacenter.  I 
rebooted many weeks after, and the machine didn't come back up.  So I 
went to the datacenter tonight to find out that the server was stuck 
at "GRUB". Nothing more.
So I decided to re-install Centos4.  The partitions were 
already there, so I just made sure partitions were formatted and the 
install went fine.  When I rebooted, "GRUB".


I'm at this point and I don't know exactly what I could do to 
diagnose furthere.


Any ideas?


grub cannot find its second stage. Are you booting from a mirrored 
partition?


Yes

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Stuck at GRUB

[Ugo Bellavance]

Hi,

I have a Dell poweredge 400SC here with an LSI scsi card.

I installed Centos4 a while ago and put it in a datacenter.  I rebooted 
many weeks after, and the machine didn't come back up.  So I went to the 
datacenter tonight to find out that the server was stuck at "GRUB". 
Nothing more.


	So I decided to re-install Centos4.  The partitions were already there, 
so I just made sure partitions were formatted and the install went fine. 
 When I rebooted, "GRUB".


	I'm at this point and I don't know exactly what I could do to diagnose 
furthere.


Any ideas?

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] RAID, LVM, extra disks...

[Ugo Bellavance]

Hi,

This is my current config:

/dev/md0 -> 200 MB -> sda1 + sdd1 -> /boot
/dev/md1 ->  36 GB -> sda2 + sdd2 -> form VolGroup00 with md2
/dev/md2 ->  18 GB -> sdb1 + sde1 -> form VolGroup00 with md1

sda,sdd -> 36 GB 10k SCSI HDDs
sdb,sde -> 18 GB 10k SCSI HDDs

I have added 2 36 GB 10K SCSI drives in it, they are detected as sdc and 
sdf.


What should I do if I want to optimize disk space?

The simplest solution would be to create /dev/md3 out of sdc1 and sdf1, 
and add it to the VG, and increase the size of my /vz logical volume.


However, if I could convert that to a RAID5 (it could be possible to 
re-install, but I would rather not), I could have 6 drives in RAID5, so 
I'd have 5x36 GB (180) of space available total, instead of 3*36 (108).


Also, please note that this is a Dual PIII 1.2 Tualatin. I'm runing 
OpenVZ with virtual machines that are not really CPU expensive, but I 
would not like to see my processors spending most of their time 
computing XORs for the RAID5.


Any suggestions/tips welcome.

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] lvresize --resizefs

[Ugo Bellavance]

Hi,

	There is a difference between the help of lvresize and its man page. 
In the manpage, there is nothing about the -r or --resizefs function. 
Centos4.


[EMAIL PROTECTED] ~]# lvresize
  Please specify either size or extents (not both)
  lvresize: Resize a logical volume

lvresize
[-A|--autobackup y|n]
[--alloc AllocationPolicy]
[-d|--debug]
[-h|--help]
[-i|--stripes Stripes [-I|--stripesize StripeSize]]
{-l|--extents [+|-]LogicalExtentsNumber[%{VG|LV|FREE}] |
 -L|--size [+|-]LogicalVolumeSize[kKmMgGtTpPeE]}
[-n|--nofsck]
[-r|--resizefs]
[-t|--test]
[--type VolumeType]
[-v|--verbose]
[--version]
LogicalVolume[Path] [ PhysicalVolumePath... ]

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] kernel-2.6.9-55.0.12.EL not booting (lvm? 3ware?)

[Ugo Bellavance]

Hi,

Here's he setup.

Original Kernel:

kernel-smp-2.6.9-42.EL, booting fine.

Error with kernel-2.6.9-55.0.12.EL (and kernel-2.6.9-55.0.12.ELsmp)

No volume groups found

Volume group "VolGroup00" not found.

ERROR: /bin/lvm exited abnormally!

The system has a

3ware 9000 Storage Controller
Model: 9500S-4LP
Firmware FE9X 2.08.00.006, BIOS BE9X 2.03.01.052, Ports: 4.

Anyone seen this?

I've googled but can't find anything, except that:

[184523] -add support for 3ware 9650SE SATA-RAID (Chip Coldwell) ... in 
2.6.9-44


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: apc deemon

[Ugo Bellavance]

[EMAIL PROTECTED] wrote:

Euroka wrote:

Hi,

I'm using Centos 5 (64bit) and connected an APC battery to the server
and used the USB cable for monitoring this setup, but I can't find the
apc daemon anymore in the offcial yum repo's.

http://dag.wieers.com/rpm/packages/apcupsd/


Home page of apcupsd has changed to http://apcupsd.com/
Dag's page points to old url.


I think you should tell Dag, not the list.

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: apc deemon

[Ugo Bellavance]

Euroka wrote:

Hi,

I'm using Centos 5 (64bit) and connected an APC battery to the server 
and used the USB cable for monitoring this setup, but I can't find the 
apc daemon anymore in the offcial yum repo's.


http://dag.wieers.com/rpm/packages/apcupsd/

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: webserver setup

[Ugo Bellavance]

Andreas Kuntzagk wrote:

Hi,

I have setup a webserver on CentOS 4.5 with NameVirtualHost.
Two VirtualHost should be reachable by port 80 from outside, while the
third (default) should be only reachable by https from outside but by
http from inside. 
Since all share the same IP firewalling seems to be out of question. So

is there a way to restrict that in Apache config?

regards, Andreas


Configure your virtualhosts

a "SSLRequireSSL" directive would force it to be https only.  When using 
http, it would give "Permission denied".  Or a mod_rewrite.


Or you could also configure your http virtualhosts in httpd.conf, while 
configuring your https VH in ssl.conf.


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Migration from RH9

[Ugo Bellavance]

James A. Peltier wrote:

Leonel Nunez wrote:

I can second this.  I am in the process of migrating 5 research labs 
from Suse 10.0 to CentOS 5 (for various reasons).  The migration has 
been in testing phase for over 3 months and a lot of bugs have been 
found and corrected in that time.


A migration from any OS to another is a very tedious and time consuming 
step.  You will need to work on each part of the migration individually. 
 Start with the services that you are most familiar with or that you 
feel you could learn the quickest.


Setup a machine with CentOS 5 and begin testing that service.  When you 
are confident that said service is operating as it should, shift that 
service from the production server to your testing server.  Let it run 
there for a bit because chances are you'll find bugs and that will give 
you a chance to fail the service back over (if necessary) while you 
correct the issue.


Once you've gotten all the services over to the new box you'll be happy 
to know you did it the "right way" and that you've incurred the least 
amount of pain for you, your fellow workers who work with you and your 
users.


IMHO, you should spend a lot of time testing the Perforce migration, 
followed by your web services.  Migration of any SCM is a potentially 
complicated operation.  I haven't used Perforce before, but be careful.


Secondly, careful testing of your web services is crucial.  You'll most 
likely be upgrading version of Apache, PHP and libraries at the same 
time which can break things like backward compatibility. Samba depending 
on it's function within your institution would be a close third, if not 
a tie for number 2, but that's up to you.


The squid services are probably not all that complicated if they're only 
using a caching server (forward or reverse).


Of course, with proper software unit testing and a bit of elbow grease 
I'm sure it will all go over well.


There are various papers on best practices for OS migrations and various 
other system administrator task on the web just google for migration 
best practices and you'll find lots.




Wow, excellent advice!

Thanks!

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Migration from RH9

[Ugo Bellavance]

Craig White wrote:

On Thu, 2007-11-22 at 09:51 -0500, Ugo Bellavance wrote:





questions were far to general to get specific responses


That is what I thought.



There is no supported upgrade path for either RHL 9 or Fedora 4 to
CentOS but you certainly can use anaconda to do that by appending...

linux upgradeany


In fact I won't upgrade the current machines, as they're old.  I'd setup 
new servers and migrate services one to one from the old to the new ones.



You get to keep the pieces that are broken.

per info on Centos web site...
http://centos.org/modules/smartfaq/faq.php?faqid=43

CentOS 3.x is in maintenance updates only mode...hardly makes sense to
me to install this on anything UNLESS you have specific software that
will only run on 2.4 kernels

CentOS 4.x enters into maintenance updates only mode March 1st, 2008 -
not that far off

I don't know anything about 'perforce environment' and thus have nothing
to offer. Similarly, I haven't use squid for many years and likewise
dansguardian.


Perforce is a CVS-like system (proprietary).


Apache/PHP/MySQL are all starting to show their age on CentOS-4.x but
thankfully, there is CentOS Plus repository for updated php & mysql but
if I were thinking in terms of updates that I didn't have to bother with
for a few years, I would certainly go all the way to CentOS 5

You might find some useful information here...

http://centos.org/search.php?query=upgradeany&andor=AND&action=results&submit=Search


All very good info, thanks!

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Migration from RH9

[Ugo Bellavance]

Ugo Bellavance wrote:

Hi,

I will probably have to migrate a few servers running Red Hat 9 to 
CentOS and I'd like to have opinions.  One server to migrate is running 
Fedora core 4.


Anyone has experience with that?  I'm not too scared about sendmail, 
Bind, apache/PHP (except that maybe the PHP upgrade may break the 
website, but this is easily managed and can be tested), ftp.


However, I don't know dansguardian much, but I guess it would be a good 
idea to integrate it with squid.


I don't know squid much, but depending on what they're doing with it, it 
should be fairly simple.


For samba, are there major changes between the default version in RH9 
and Centos4 or 5?


Anyone migrated a perforce environment?

I'm also wondering whether I migrate them to centos3, 4, or 5.  3 would 
be to make the transition easier, while still letting them have many 
years of security updates.


4 is the version I use most, I barely touched 5.

5 would be for xen and... your ideas?

I usually use OpenVZ for virtualization, but they might want to 
virtualize windows servers too, which OpenVZ can't do.


Any input would be appreciated,


No answer yet... I guess I must have done something wrong in my post.

I'll try to get more details about what is to be done and I'll post 
again.  Still, if you have an advice about anything in my original post, 
 please let me know.


Regards,
Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: HDD usage oddity

[Ugo Bellavance]

Stephen Harris wrote:

On Wed, Nov 21, 2007 at 07:37:31PM -0500, Ugo Bellavance wrote:

2 of my Centos4 servers are showing weird behavior.

[EMAIL PROTECTED] ~]# du -sh /var/
1.8G/var/
[EMAIL PROTECTED] ~]# df -h /var
FilesystemSize  Used Avail Use% Mounted on
/dev/sda7 4.3G  3.9G  181M  96% /var


There are two things to consider.

ONE
===

If you have large block sizes and lots and lots of small files then
"du" will report a number quite a lot smaller than "df".  "df" reports
on actually amount free; "du" reports on apparent size of the files used.

TWO
===

If you delete a file that is still "open" (typically a log file, or
a file being written to in /var/tmp) then Unix doesn't actually delete
the file at that point; it removes the directory entry and marks the file
to be deleted when no one is using it any more.

I've seen developers create debug logs in /tmp that grew to 2Gb then
delete them to save space... while the program was still writing more
debug messages, so they didn't save any space at all.



That was it...  I deleted a log that was still opened by syslog, and 
other by snort.


After restarting snort and syslog, I got the right reading.

diskcheck will not stop bugging me :)

Thanks a lot!

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Backups on external USB HDD

[Ugo Bellavance]

Ugo Bellavance wrote:

David Ferreira wrote:

Hi

If Device 009 is your USB HD, then, your 
/etc/hotplug/usb/usbhd.usermap sould look like that:


#usb module   match_flags idVendor idProduct bcdDevice_lo bcdDevice_hi 
bDeviceClass bDeviceSubClass bDeviceProtocol bInterfaceClass 
bInterfaceSubClass bInterfaceProtocol driver_info
usbbackup0x00030x04fc0x0c250x0x0x00
0x000x000x000x000x000x



Well, now the script. only one question, Do you use backup exec to 
write on linuxbox fs and then, rsync to the usb hd?, or do you want to 
backup exec directly backup to the usb hd? If you do a rsync from 
local linux fs to the USB disk, you can umount the fs in the script 
when it's finish. If you choose the second option (Backup diretly to 
the usb hd) you sould find other way to umount the usb drive once you 
have finished backup (I think the easiest  way is a cron job). I think 
first option is beter, so you can use this pseudo-script in 
/etc/hotplug/usb/usbhd:


#!bin/bash
mount -t auto /dev/sd0 /mnt/backup
rsync -Pavz --delete /tmp/backup /mnt/backup
umount /mnt/backup

If you prefere second option, you must made accesible /mnt/backup as 
you prefer to your windows box, quit rsync and umount from script, and 
create a cron job to umount the usb drive.

I think now is enough clear what do you have to do.


Yes, I'd rather use the 2nd option, as the USB HDDs' size is 250 GB, 
while I only have about 100 GB locally.


Therefore, I only need the first line, so that it is automatically 
mounted?  In fact, what I want is to always have /dev/sda1 mounted as 
/mnt/usbbackup (example) as soon as it is connected.


My backup usually takes about 5 hours, so if I start it at 19.00, it 
will be safe to umount the volume anytime after 4AM, for example, using 
a cronjob.


Does this make sense?

Is there a way to test it w/o being on site (one HDD is already connected).


The mount didn't work finally.  I didn't have much time to test and I'm 
not on-site, so I decided so simply do a 'mount /dev/sda1 
/mnt/usbbackup;service smb start' at 18:00 and the opposite at 3AM.


I start my backup job at 20:00 so I should be fine.

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] HDD usage oddity

[Ugo Bellavance]

Hi,

2 of my Centos4 servers are showing weird behavior.

[EMAIL PROTECTED] ~]# du -sh /var/
1.8G/var/
[EMAIL PROTECTED] ~]# df -h /var
FilesystemSize  Used Avail Use% Mounted on
/dev/sda7 4.3G  3.9G  181M  96% /var

[EMAIL PROTECTED] var]# df -h /var
FilesystemSize  Used Avail Use% Mounted on
/dev/sda7 2.9G  2.7G 0 100% /var
[EMAIL PROTECTED] var]# du -sh /var/
1.6G/var/

du seems to be reporting the right size

Anyone seeing this?

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: how to Load balancing

[Ugo Bellavance]

D. Bettancourt M. wrote:


Hi

I need informmation about this Topic: Load balancing,,

I have 2 nic to internet, 1 nic to DMZ, and 1 nic to LAN, but I don,t 
know how do that.


Where I can found information ??


What do you need to load-balance? Internet connections or services on 
your server?


Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: "su" error need help???

[Ugo Bellavance]

mcclnx mcc wrote:
I 'su" to root and run command.  I got error message. 
Anyone know how to fix it?


=

/home/app/oracle/product/oms10g/install/unix/scripts/omsstup
start
/bin/su: user %s_userName% does not exist



I guess the user that is called doesn't exist.

You'd have to look in the script to see which user is called there.

Was there an installation procedure or directives?


___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Migration from RH9

[Ugo Bellavance]

Hi,

	I will probably have to migrate a few servers running Red Hat 9 to 
CentOS and I'd like to have opinions.  One server to migrate is running 
Fedora core 4.


Reason for migrating to CentOS: no more security updates available, 
hardware change planned anyway.


Here are the services to be migrated:

sendmail
pop (not sure which daemon, I think it is running the default daemon in RH9)
DNS
apache+PHP website
FTP
Dansguardian
Samba
Perforce
Squid

I haven't seen any directories like LDAP or windows Active Directory, so 
I assume for now that all authentication is done locally.


Anyone has experience with that?  I'm not too scared about sendmail, 
Bind, apache/PHP (except that maybe the PHP upgrade may break the 
website, but this is easily managed and can be tested), ftp.


However, I don't know dansguardian much, but I guess it would be a good 
idea to integrate it with squid.


I don't know squid much, but depending on what they're doing with it, it 
should be fairly simple.


For samba, are there major changes between the default version in RH9 
and Centos4 or 5?


Anyone migrated a perforce environment?

I'm also wondering whether I migrate them to centos3, 4, or 5.  3 would 
be to make the transition easier, while still letting them have many 
years of security updates.


4 is the version I use most, I barely touched 5.

5 would be for xen and... your ideas?

I usually use OpenVZ for virtualization, but they might want to 
virtualize windows servers too, which OpenVZ can't do.


Any input would be appreciated,

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Backups on external USB HDD

[Ugo Bellavance]

David Ferreira wrote:

Hi

If Device 009 is your USB HD, then, your /etc/hotplug/usb/usbhd.usermap 
sould look like that:


#usb module   match_flags idVendor idProduct bcdDevice_lo bcdDevice_hi 
bDeviceClass bDeviceSubClass bDeviceProtocol bInterfaceClass 
bInterfaceSubClass bInterfaceProtocol driver_info
usbbackup0x00030x04fc0x0c250x0x0x00
0x000x000x000x000x000x



Well, now the script. only one question, Do you use backup exec to write 
on linuxbox fs and then, rsync to the usb hd?, or do you want to backup 
exec directly backup to the usb hd? If you do a rsync from local linux 
fs to the USB disk, you can umount the fs in the script when it's 
finish. If you choose the second option (Backup diretly to the usb hd) 
you sould find other way to umount the usb drive once you have finished 
backup (I think the easiest  way is a cron job). I think first option is 
beter, so you can use this pseudo-script in /etc/hotplug/usb/usbhd:


#!bin/bash
mount -t auto /dev/sd0 /mnt/backup
rsync -Pavz --delete /tmp/backup /mnt/backup
umount /mnt/backup

If you prefere second option, you must made accesible /mnt/backup as you 
prefer to your windows box, quit rsync and umount from script, and 
create a cron job to umount the usb drive.

I think now is enough clear what do you have to do.


Yes, I'd rather use the 2nd option, as the USB HDDs' size is 250 GB, 
while I only have about 100 GB locally.


Therefore, I only need the first line, so that it is automatically 
mounted?  In fact, what I want is to always have /dev/sda1 mounted as 
/mnt/usbbackup (example) as soon as it is connected.


My backup usually takes about 5 hours, so if I start it at 19.00, it 
will be safe to umount the volume anytime after 4AM, for example, using 
a cronjob.


Does this make sense?

Is there a way to test it w/o being on site (one HDD is already connected).

Regards,
Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Backups on external USB HDD

[Ugo Bellavance]

Chris wrote:

On Tue, 20 Nov 2007 06:43:54 -0500
Ugo Bellavance <[EMAIL PROTECTED]> wrote:


Hi,

I'm trying to implement this:

I have:

- A windows 2000 server
- A centos 4 server
- Monday, Tuesday, Wednesday, Thursday, Friday, Monthly.

The tape drive in the windows server died recently and I decided to 
switch to USB external drives.  However, the USB controller in the 
windows server is only 1.1, so it is very slow.


I didn't want to install a 2.0 USB controller in the windows server 
since it is a brand-name and I didn't want to make it unstable, so I 
decided to make the backups of the windows server (using Backup Exec)

on the linux box.

So backup exec writes on the linux box via samba, directly on the USB 
drive. (I thought of writing on the linux box FS directly, then 
rsync'ing to the USB drive), but the space available on the local FS

is about 90 gigs while the external USB drives are 250 gigs, meaning
that I can keep like 4 weeks of data on the USB drives (using backup
exec settings), while I could only keep 1 or 2 weeks otherwise.

My questions is:

-how can the USB drives be umounted/mounted automatically when the 
person on site changes it (monday to tuesday, for example).  There

will always be only one HDD conected at the time.



I can't answer on the auto mounting however, consider this -
Run something like BackupPC on the Cent box. That in turn allows rsync
to fetch from the Windows Server without the extra overhead of samba
etc.


Ok, but we already have Backup Exec on the server, that can backup 
exchange, brick-level and full information store, system state, etc.


I'm just looking for a way to be able to have backupExec to write to the 
external drive.


Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Backups on external USB HDD

[Ugo Bellavance]

David Ferreira wrote:
I've used simple  scripts for similar backup system. I used hotplug 
facilities and rsync to avoid copying inecesary data.

You must see the USB id which lsusb

in /etc/hotplug/usb/ create the folowing two files:
usbhd.usermap containing the folowing line:
usbbackup 0x0003 0x05e3 0x0702 0x 0x 0x00 0x00 0x00 0x00 0x00 
0x00 0x


I don't need the script, I just need the windows box to be able to write 
 to the external HDD, and that the external HDD can be changed (not 
when writing, of course) without any manual intervention.


This is the output of lsusb

[EMAIL PROTECTED]
Unknown line at line 5959
Unknown line at line 5960
Unknown line at line 5961
Unknown line at line 5962
Unknown line at line 5963
Unknown line at line 5964
Unknown line at line 5965
Unknown line at line 5966
Unknown line at line 5967
Unknown line at line 5968
Unknown line at line 5969
Unknown line at line 5970
Bus 004 Device 001: ID :
Bus 003 Device 001: ID :
Bus 002 Device 001: ID :
Bus 001 Device 009: ID 04fc:0c25 Sunplus Technology Co., Ltd
Bus 001 Device 001: ID :

The device is most likely 001/009 (Sunplus).

lsusb -v shows this (stripped output):

Bus 001 Device 009: ID 04fc:0c25 Sunplus Technology Co., Ltd
Device Descriptor:
  bLength18
  bDescriptorType 1
  bcdUSB   2.00
  bDeviceClass0 (Defined at Interface level)
  bDeviceSubClass 0
  bDeviceProtocol 0
  bMaxPacketSize064
  idVendor   0x04fc Sunplus Technology Co., Ltd
  idProduct  0x0c25
  bcdDevice1.03
  iManufacturer   2 Sunplus Technology Inc.
  iProduct3 USB to Serial-ATA bridge
  iSerial 1 WDC WD2500 WD-WXE807734118
  bNumConfigurations  1

How would I build my usbhd.usermap file?

Regards,
Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Backups on external USB HDD

[Ugo Bellavance]

Hi,

I'm trying to implement this:

I have:

- A windows 2000 server
- A centos 4 server
- Monday, Tuesday, Wednesday, Thursday, Friday, Monthly.

The tape drive in the windows server died recently and I decided to 
switch to USB external drives.  However, the USB controller in the 
windows server is only 1.1, so it is very slow.


I didn't want to install a 2.0 USB controller in the windows server 
since it is a brand-name and I didn't want to make it unstable, so I 
decided to make the backups of the windows server (using Backup Exec) on 
the linux box.


So backup exec writes on the linux box via samba, directly on the USB 
drive. (I thought of writing on the linux box FS directly, then 
rsync'ing to the USB drive), but the space available on the local FS is 
about 90 gigs while the external USB drives are 250 gigs, meaning that I 
can keep like 4 weeks of data on the USB drives (using backup exec 
settings), while I could only keep 1 or 2 weeks otherwise.


My questions is:

-how can the USB drives be umounted/mounted automatically when the 
person on site changes it (monday to tuesday, for example).  There will 
always be only one HDD conected at the time.


Thanks,
Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] FTP with web interface

[Ugo Bellavance]

Hi,

I'm looking for a simple way to be able to administer ftp 
accounts on Centos4 (or 5, if needed).  The best would be to have a web 
interface to do so, but I think I'd rather avoid using webmin...

I've seen some tutorials for getting ftp servers to work with MySQL DBs, 
but I'd like to know how people are working.  I'd like to have a separate 
set of users for ftp than for system.

Thanks,
-- 
Ugo Bellavance

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Linux PDA/smartphone sync?

[Ugo Bellavance]

Hi,

I think I'll finally be able to move away from windows... I just
installed linux and I'm wondering what people use as smartphones/PDA.

I currently use a motorola V710 phone, which allows me to sync my
contacts and calendar, which is all I need.  I may change it soon, so I'd
like to know what is best to work with linux (I use evolution right now,
connected to an exchange (eventually scalix) server).

I'm trying to get moto4linux to work, but I'm not even sure if 
I'll be able to sync my calendar + contacts with this.

Any advice?

Regards,
-- 
Ugo Bellavance

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Best laptop for CentOS

[Ugo Bellavance]

Ern jura wrote:

Which is the best laptop to run centos on and where can I buy one
without  Windows Vista pre loaded cause I will uninstall it any way.
Please tell me the best sites that ship worldwide.


From the linux journal review of this month:

http://www.emperorlinux.com/

Lenovo ThinkPad T61 - linux version

www.lenovo.com

R cubed

www.shoprcubed.com

HPC systems

www.hpcsystems.com

Dell Inspiron 1420N

www.dell.com/linux

Also check:

www.compamerica.com
www.polywell.com
www.thinkmate.com

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: FW: Logwatch for XXXXXXX.kd4efm.org (Linux)

[Ugo Bellavance]

Evans F. Mitchell KD4EFM / AFA2TH / WQFK-894 wrote:

Found an error or two from my logwatch report from yesterday,
thought I would share this in hopes this is just first time
run of the problem I noticed in the Kernel report section...

Also not sure why there's an issue with automount either
but I guess I could ask on that issue as well.

I am not worried about the NAMED error, this is something that
happens due to one of the services that is installed on the box,
as it is HAM RADIO related only.

Any feedback? I will be looking for it... 


some items will be X'ed for protection reasons.



 WARNING:  Kernel Errors Present
end_request: I/O error, dev fd0, sector ...:  2 Time(s)


You tried to read or write to a floppy and it kind of failed...

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: slow SMTP-AUTH

[Ugo Bellavance]

Ray Van Dolson wrote:

On Wed, Jul 25, 2007 at 07:35:58PM -0400, Ugo Bellavance wrote:

Ugo Bellavance wrote:

Brett Schroeder wrote:

Ugo Bellavance wrote:

Hi,

   I installed sendmail on a centos-4 openvz VM.  I set up SMTP-AUTH
and when I try to send mail using this server, there is a big delay in
which my e-mail client (Mozilla Thunderbird) just hangs.  The server, on
the other hand, is idle at the same time.  It does that no matter if I
use no encryption, TLS, or SSL.  When the client is 'hung' sendmail, on
the server side haven't logged anything yet for this connection.

Any idea?
Since nothing is logged on the server, it sounds like there is a 
firewall (or
other networking problem) in the way. You should use WireShark (or 
tcpdump) on

the client to capture the network traffic.
I disabled the firewall on the server, no luck.  I tried sniffing, I see 
nothing weird.


  5.512292 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [PSH, 
ACK] Seq=350 Ack=10923 Win=17461 Len=53
  5.512984 72.55.149.51 -> 74.59.221.180 TCP submission > 17274 [PSH, 
ACK] Seq=10923 Ack=403 Win=7504 Len=277
  5.724172 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [ACK] 
Seq=403 Ack=11200 Win=17184 Len=0
 22.884013 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [PSH, 
ACK] Seq=403 Ack=11200 Win=17184 Len=53
 22.884918 72.55.149.51 -> 74.59.221.180 TCP submission > 17274 [PSH, 
ACK] Seq=11200 Ack=456 Win=7504 Len=85

 22.970714 74.59.221.180 -> 72.55.149.51 TLSv1 Application Data
 22.971120 72.55.149.51 -> 74.59.221.180 TLSv1 Application Data



Can you run strace on sendmail during a connection and see what it's
doing?  DNS is often the culprit when delays are concerned, so look for
anything along those lines.

Ray


Here is an interesting part, but I can't tell why it took 2.292863 
seconds to do that:


09:07:47 time(NULL) = 1185455267 <0.07>
09:07:47 rt_sigprocmask(SIG_UNBLOCK, [ALRM], [ALRM], 8) = 0 <0.08>
09:07:47 select(8, [4 5 6 7], NULL, NULL, {5, 0}) = 1 (in [6], left {2, 
707000}) <2.292863>
09:07:49 accept(6, {sa_family=AF_INET, sin_port=htons(25919), 
sin_addr=inet_addr("74.59.221.180")}, [16]) = 9 <0.23>

09:07:49 rt_sigprocmask(SIG_BLOCK, [ALRM], [], 8) = 0 <0.10>
09:07:49 time(NULL) = 1185455269 <0.07>
09:07:49 pipe([10, 11]) = 0 <0.21>
09:07:49 rt_sigprocmask(SIG_BLOCK, [CHLD], [ALRM], 8) = 0 <0.08>
09:07:49 clone(Process 16312 attached

Some operations are unfinished like this one:

[pid 31853] 09:07:49 select(8, [4 5 6 7], NULL, NULL, {5, 0} ...>



There is a 5-second gap here (from one PID to the other, weird):

[pid 16312] 09:07:49 time(NULL) = 1185455269 <0.10>
[pid 16312] 09:07:49 read(9, 0x8968260, 5) = -1 EAGAIN (Resource 
temporarily unavailable) <0.12>

[pid 16312] 09:07:49 time(NULL) = 1185455269 <0.09>
[pid 16312] 09:07:49 select(10, [9], NULL, [9], {3600, 0} 
[pid 31853] 09:07:54 <... select resumed> ) = 0 (Timeout) <4.999631>
[pid 31853] 09:07:54 rt_sigprocmask(SIG_BLOCK, [ALRM], [], 8) = 0 <0.08>
[pid 31853] 09:07:54 time(NULL) = 1185455274 <0.08>
[pid 31853] 09:07:54 open("/proc/loadavg", O_RDONLY) = 9 <0.37>
[pid 31853] 09:07:54 fstat64(9, {st_mode=S_IFREG|0444, st_size=0, ...}) 
= 0 <0.08>
[pid 31853] 09:07:54 read(9, "0.01 0.02 0.00 3/49 16395\n", 1024) = 26 
<0.15>

[pid 31853] 09:07:54 close(9)   = 0 <0.13>
[pid 31853] 09:07:54 time(NULL) = 1185455274 <0.07>
[pid 31853] 09:07:54 open("/proc/loadavg", O_RDONLY) = 9 <0.15>




___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Cricket

[Ugo Bellavance]

centos wrote:

Thank you Ugo,

But I was asking about Cricket and not Cacti.


You're right I confused them...


please let me know if you have installed Cricket as well.


No, I can't even remember what it does.

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: slow SMTP-AUTH

[Ugo Bellavance]

Ugo Bellavance wrote:

Brett Schroeder wrote:

Ugo Bellavance wrote:

Hi,

I installed sendmail on a centos-4 openvz VM.  I set up SMTP-AUTH
and when I try to send mail using this server, there is a big delay in
which my e-mail client (Mozilla Thunderbird) just hangs.  The server, on
the other hand, is idle at the same time.  It does that no matter if I
use no encryption, TLS, or SSL.  When the client is 'hung' sendmail, on
the server side haven't logged anything yet for this connection.

Any idea?


Since nothing is logged on the server, it sounds like there is a 
firewall (or
other networking problem) in the way. You should use WireShark (or 
tcpdump) on

the client to capture the network traffic.


I disabled the firewall on the server, no luck.  I tried sniffing, I see 
nothing weird.


  5.512292 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [PSH, 
ACK] Seq=350 Ack=10923 Win=17461 Len=53
  5.512984 72.55.149.51 -> 74.59.221.180 TCP submission > 17274 [PSH, 
ACK] Seq=10923 Ack=403 Win=7504 Len=277
  5.724172 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [ACK] 
Seq=403 Ack=11200 Win=17184 Len=0
 22.884013 74.59.221.180 -> 72.55.149.51 TCP 17274 > submission [PSH, 
ACK] Seq=403 Ack=11200 Win=17184 Len=53
 22.884918 72.55.149.51 -> 74.59.221.180 TCP submission > 17274 [PSH, 
ACK] Seq=11200 Ack=456 Win=7504 Len=85

 22.970714 74.59.221.180 -> 72.55.149.51 TLSv1 Application Data
 22.971120 72.55.149.51 -> 74.59.221.180 TLSv1 Application Data


(I removed all lines containing SSH or ssh).

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: slow SMTP-AUTH

[Ugo Bellavance]

Brett Schroeder wrote:

Ugo Bellavance wrote:

Hi,

I installed sendmail on a centos-4 openvz VM.  I set up SMTP-AUTH
and when I try to send mail using this server, there is a big delay in
which my e-mail client (Mozilla Thunderbird) just hangs.  The server, on
the other hand, is idle at the same time.  It does that no matter if I
use no encryption, TLS, or SSL.  When the client is 'hung' sendmail, on
the server side haven't logged anything yet for this connection.

Any idea?


Since nothing is logged on the server, it sounds like there is a firewall (or
other networking problem) in the way. You should use WireShark (or tcpdump) on
the client to capture the network traffic.


Hmmm, interesting... IDENT, maybe?

I'll look into that.

Thanks,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: Cricket

[Ugo Bellavance]

Centos wrote:

Hello

Has any one installed Cricket on Cenos ?
it seems there is a patch that should apply on Format.pm to specify 
Achitecture,

but I can not find it .

Thanks


I wrote an how-to on the centos wiki on it.  Make sure you use PHP5.

http://wiki.centos.org/HowTos/Cacti_on_CentOS_4.x

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] slow SMTP-AUTH

[Ugo Bellavance]

Hi,

	I installed sendmail on a centos-4 openvz VM.  I set up SMTP-AUTH and 
when I try to send mail using this server, there is a big delay in which 
my e-mail client (Mozilla Thunderbird) just hangs.  The server, on the 
other hand, is idle at the same time.  It does that no matter if I use 
no encryption, TLS, or SSL.  When the client is 'hung' sendmail, on the 
server side haven't logged anything yet for this connection.


Any idea?

Regards,

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: SSH remote port forward

[Ugo Bellavance]

Tony Mountifield wrote:

In article <[EMAIL PROTECTED]>, Ugo Bellavance <[EMAIL PROTECTED]> wrote:

Hi,

	I'm trying to do a remote port forward, and the remote hosts only 
listens to 127.0.0.1 on the forwarded port.


You can't control that from the client end, as it could be a security risk
for the remote installation.


Makes sense.


If the remote host is under your control and is running sshd from OpenSSH,
you can add the following directive to /etc/ssh/sshd_config on that system:

GatewayPorts yes

That will tell the sshd to make forwarded ports listen on 0.0.0.0 instead
of on 127.0.0.1


Excellent, thanks for the help it now works ! :).

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] SSH remote port forward

[Ugo Bellavance]

Hi,

	I'm trying to do a remote port forward, and the remote hosts only 
listens to 127.0.0.1 on the forwarded port.


[EMAIL PROTECTED] ~]$ netstat -ln
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address   Foreign Address 
State
tcp0  0 127.0.0.1:5500  0.0.0.0:* 
LISTEN
tcp0  0 0.0.0.0:56660.0.0.0:* 
LISTEN
tcp0  0 ::1:5500:::* 
LISTEN
tcp0  0 :::22   :::* 
LISTEN


(I'm trying to tunnel a vnc client connection.

From putty's logs, you can see that the tunnel is requested:

2007-06-12 17:34:15 Opened channel for session
2007-06-12 17:34:15 Requesting remote port 5500 forward to 10.0.0.1:5500
2007-06-12 17:34:15 Remote port forwarding from 5500 enabled

(I changed the real IP address to 10.0.0.1).

In putty, the "Local ports accept connections from other hosts" and 
"Remote ports do the same" are checked.


Any ideas?

Regards,

ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

[CentOS] Re: kernel-smp not upgrading

[Ugo Bellavance]

Scott Silva wrote:

Ugo Bellavance spake the following on 6/11/2007 9:44 PM:

Hi,

Box is a dual-dual core opteron.  Centos4.  Just upgraded t the
latest kernel (2.6.9-55).

The UP kernel installed fine, but the smp kernel didn't install. 
What should I do?  Is that normal? Original kernel was a smp.


Regards,

Ugo

I had that problem when the updates came out on one of my servers. Manually
installing the kernel again seemed to fix it. It actually started the install,
and the rpm database said it was installed, but the kernel file and the initrd
weren't anywhere to be found. But the rest of the kernel related files were
there in boot.


Same thing here,  fixed the same way :).

Ugo

___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos