OpenSolaris derived ZFS NAS/ SAN (OmniOS, OpenIndiana, Solaris and napp-it)

mrjester

n00b
Joined
May 16, 2014
Messages
4
update: 17.May
I have removed a check for valid ip4 adresses in admin.pl.
You may reload the 0.9 f1 preview from today.

I have not checked myself as my net is currently ip4 only.

Your changes worked. Thank you.
Is it possible to have multiple ranges listed? Specifically, I have a need for 1 v6 subnet and 2x v4 subnets.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
Your changes worked. Thank you.
Is it possible to have multiple ranges listed? Specifically, I have a need for 1 v6 subnet and 2x v4 subnets.

Seems quite usefull without the need to use the firewall.
Please re-download the f1 preview.

You can now enter a commalist with allowed ip adresses (or the beginning of allowed addresses)
 

mrjester

n00b
Joined
May 16, 2014
Messages
4
Seems quite usefull without the need to use the firewall.
Please re-download the f1 preview.

You can now enter a commalist with allowed ip adresses (or the beginning of allowed addresses)

Awesome. Thank you. Tested successful as well.
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
I have a disk that return some error message:

uLkVrQ.png

c4t5000CCA22EC1B9A2d0 smart_health is "FAILED!"

What to do? I tried to locate the faulty disk by unplugging the disks one by one but it seems it take a while before they are marked as "REMOVED".

Is it serious? Do you think i should replace the drive?

Do you think the disks are overheating? (They reached around 49/52°C maximum before i blocked the air leaks in the case with some foam, after that little mod the temps dropped to the ones showed on the first screenshot)
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
Also about shares, is it possible to share a few folders (for guest users, without needing a pw) located in a dataset that also contain hidden/private folders? Or do i need to make a dataset with dedicated sharing rules set for each folder?

Like:
dataset/directory1 (shared for everyone, read only)
dataset/directory2 (hidden and private)

Or:
dataset1/all the folders inside (guest shared/read only)
dataset2/all the folders inside (root only)
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
Also about shares, is it possible to share a few folders (for guest users, without needing a pw) located in a dataset that also contain hidden/private folders? Or do i need to make a dataset with dedicated sharing rules set for each folder?

Like:
dataset/directory1 (shared for everyone, read only)
dataset/directory2 (hidden and private)

Or:
dataset1/all the folders inside (guest shared/read only)
dataset2/all the folders inside (root only)

If you share a filesystem with guest=enable, there is no user authentication where you can restrict acess - > all or nothing
If you share with guest=disable, you always need a login so you can use ACL to restrict access based on a user or everyone@
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
c4t5000CCA22EC1B9A2d0 smart_health is "FAILED!"

What to do? I tried to locate the faulty disk by unplugging the disks one by one but it seems it take a while before they are marked as "REMOVED".

Is it serious? Do you think i should replace the drive?

Do you think the disks are overheating? (They reached around 49/52°C maximum before i blocked the air leaks in the case with some foam, after that little mod the temps dropped to the ones showed on the first screenshot)

Temp is ok

what you can do:
- click on smart serial to read smartlog, check homepage of smartmonttols about how to read
- click on smart_check for a second check
- check the disk on a pc with a low level tool from the manufacturer (i would do this)
- wait until it is a real ZFS failure (smart is more a estimate)

- be prepared for real disk failures (you should always), use hot spares
do not panic about smart. Many failures come without a prior smart warning

Last word: You should always know which disk is in which slot.
Print out disk-details page and insert WWN-> slot slot informations.

As you have a monitor evalkey, you can use menu Disks-SAS2 extension where you get information about the slot where a disk is inserted and the
sas/target number that is needed to identify a disk from system-log on problems.
Print out this page and place in on top of your storage. You will need this infos one day or another:

sas2_slots.png
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
If you share a filesystem with guest=enable, there is no user authentication where you can restrict acess - > all or nothing
If you share with guest=disable, you always need a login so you can use ACL to restrict access based on a user or everyone@

But if i disable guest access to restrict access, can i still share some folders without password? (and keep all data in the same dataset)

Thanks for the smart info!
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
So, as I was saying, I'm building Zabbix templates for systems I maintain and want to monitor. Zabbix provides a basic template for Solaris, but it only monitors amount of free/used space, memory consumption, CPU usage, network usage.


Just to inform
In current napp-it you find 60s graphs of iostat for pool or single disks values like cpu%, b%, w% or iops values like read, write or wait.

Longterm graphs (hour, day, week, month are on the way)

ex: start a scrub (nice to see the 5s write buffer - small random write to large seqential write conversion - compare to sync=always)
shortterm.png
 
Last edited:

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
no, not possible.
If you disable guest you always need a login

So if i understand well, the only way to have password-free shares and restrictions on some folders is to make two datasets? (public dataset with guest and private dataset without)

edit: even using windows to make the shares from root? How does shares from nappit and shares from windows co-exist?

Temp is ok
What maximum disk temp do you recommend?
 
Last edited:

levak

Limp Gawd
Joined
Mar 27, 2011
Messages
386
Just to inform
In current napp-it you find 60s graphs of iostat for pool or single disks values like cpu%, b%, w% or iops values like read, write or wait.

Longterm graphs (hour, day, week, month are on the way)

I saw that on the printscreen you posted earlier. Very useful tool, specially for debugging. But in my case, I want more (alerts, option to select the span I want my graph to show,...). In Zabbix, I can get graphs from 10:10-10:20 for Monday 14 days ago, not just daily average, like with RRD. I'm not sure on what technology you are basing your solution, so I don't know if you have that option or not.

Any to me, it's nice to have all in one place: routers, SAN, servers,... It gives you a nice overview. But I'm sure, a lot of people, including me, will love your solution and will come handy when searching for troubles.

Matej
 

TCM2

Gawd
Joined
Oct 17, 2013
Messages
572
It should be trivial to feed data points into tools like Graphite and let those handle the statistics.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
There is a error message about mdb when trying to determine target numbers.
What OS release?

It is tested with OmniOS and LSI HBAs
 

layerbreak

n00b
Joined
Dec 13, 2011
Messages
6
was passiert denn bei Eingabe dieses Kommandos als Root:
echo "::mptsas -t" | sudo mdb -k

Die Erkennung geht nur bei Platten am mptsas Treiber (LSI HBA im IT mode)

The output:
Code:
OmniOS 5.11     omnios-6de5e81  2013.11.27
root@ripley:~# echo "::mptsas -t" | sudo mdb -k
        mptsas_t inst ncmds suspend  power
================================================================================
ffffff04f20f0000    0     0       0OFF=D3

                 The SCSI target information
mdb: couldn't read hashdata: no mapping for address

                 The smp child information
mdb: couldn't read hashtab: no mapping for address


        mptsas_t inst ncmds suspend  power
================================================================================
ffffff04eac84000    1     0       0OFF=D3

                 The SCSI target information
mdb: couldn't read hashdata: no mapping for address

                 The smp child information
mdb: couldn't read hashtab: no mapping for address

root@ripley:~#
Both LSI-HBAs are in IT-Mode.
/var/adm/Messages tells me that the disks are using mptsas driver
Example:
Code:
May 20 17:57:12 ripley genunix: [ID 408114 kern.info] /scsi_vhci/disk@g5000cca22bc8aa67 (sd19) online
May 20 17:57:12 ripley genunix: [ID 483743 kern.info] /scsi_vhci/disk@g5000cca22bc8aa67 (sd19) multipath status: degraded: path 14 mpt_sas21/disk@w5000cca22bc8aa67,0 is online
May 20 17:57:12 ripley scsi: [ID 583861 kern.info] mpt_sas16 at mpt_sas0: scsi-iport 4

Thats the output from my all-in-one and the HBA is LSI9211-8i passthrough ESXi5.5U1
Code:
root@aio:~# echo "::mptsas -t" | sudo mdb -k
mdb: couldn't read large mptsas_slots_t at 0xffffff02da015f10: only 254192 of 525792 bytes could be read
root@aio:~#
 
Last edited:

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
I have not seen this error with my machines.
Maybe you ask at the OmniOS mailing list about this output of the mdb command
 

layerbreak

n00b
Joined
Dec 13, 2011
Messages
6
I have not seen this error with my machines.
Maybe you ask at the OmniOS mailing list about this output of the mdb command

@gea
Which output do you mean? From aio or ripley?
Why I have to ask - I dont understand the input: echo "::mptsas -t" | sudo mdb -k and what you want to know.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
@gea
Which output do you mean? From aio or ripley?
Why I have to ask - I dont understand the input: echo "::mptsas -t" | sudo mdb -k and what you want to know.

This command gives you infos about the devices..
It should not end with an error.
 

natkin

n00b
Joined
Mar 31, 2014
Messages
30
I'm seeing problems with napp-it Snapshots view in Chrome since going to 0.9f1. Can anyone else confirm?

The second column has a checkbox for a header.
The third column is titled "Name" but has dates in it.
The fourth column is titled "Type" but has data size values (e.g., "2.50T")
The fifth column is titled "Creation" but has data size values.
The sixth column is titled "Used" but has data size values.
The seventh column is titled "Avail" but has the value "1.00x"
The eighth column is titled "Refer" but has the value "yes".
Columns nine through thirteen are weird too, but it seems to make sense by column fourteen, "VOLSIZE"

Kind regards.
 

dave99

2[H]4U
Joined
Jan 20, 2011
Messages
2,129
Yeah, I just looked on mine, both chrome and IE, and it's all messed up. I was on f1_2nd_preview may 14, I just updated to may 18th and it's fixed.
 
Joined
Feb 20, 2013
Messages
19
So, as I was saying, I'm building Zabbix templates for systems I maintain and want to monitor. Zabbix provides a basic template for Solaris, but it only monitors amount of free/used space, memory consumption, CPU usage, network usage.

But there have been times when someone contacted me saying: "Hey, server XX worked really slow yesterday, do you know anything about that?" This is where one could use as much information as possible and I found out disk IO helped me in many occasions.

So I created a template for disk activity monitoring. Currently I'm monitoring everything iostat shows:
- disk throughput
zabbix_transfer.jpg


- disk iops
zabbix_iops.jpg


- disk/queue busy in %:
zabbix_disk_busy.jpg


- number of active transactions
- average transaction service time
- waiting transactions (queue size)

I run iostat for 58s and save output to a file, then Zabbix agent grabs the data out of it and sends it to server. I have to create triggers as well, but I'm still thinking what threshold levels should be for triggering it.
All hard drives are auto discovered, so you just load the template into zabbix and it scans the server and add devices.

Scrub is just running on my 4x2TB raidz2 pool, that is why there is so many iops...

I still have some cleaning in template to do, then I will post a short howto here.

I've also made a SMART template, which can grep any smart status and logs it(and if there's a need, graph it).

Zabbix has it's good and bad sides. One of the good things is that you can set custom update time and amount of time it saves the data and amount of time it saves the trends(daily average). I currently have it set to 14 days, so that means I have data for every minute for the last 14 days. That's one of the benefits, compared to solutions that uses RRDs(such as Cacti, Observium(which is more userfriendly)). There's also no need to graph anything, you can call up graphs on demand and change the scale as well.

Anyone else using zabbix to monitor servers?

MAtej


Very much so, and am very interested to see the template you've made.
Been trying to get the same info through snmp, however SNMP MIBS are rather limited
 

natkin

n00b
Joined
Mar 31, 2014
Messages
30
Thank you. I wasn't aware that sub-sub-version updates were posted. I was looking for a different version in the drop-down. It would be helpful if the drop-down showed some difference in order to know without having to consult the changelog.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
Thank you. I wasn't aware that sub-sub-version updates were posted. I was looking for a different version in the drop-down. It would be helpful if the drop-down showed some difference in order to know without having to consult the changelog.

This happens especially with preview release where updates may be released every or every few days

But in general. Mini bugfixes are done within the same release number but with a new release date.
Main reson for this and not a 0.9f2 version is that you can upgrade/downgrade only last 5 releases.

I will not fill that up with such mini mini steps like fix a wrong filename or menu entry.
You can check for this mini fixes at http://napp-it.org/downloads/changelog_en.html with the release date
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
I'm still trying to identify in what slot each disk is physically located by unplugging them and write the removed disk name from the smart info page, but now the whole pool is marked as UNAVAIL because too much disk are missing (i re-plugged each disk before removing another).

All the disks are now plugged but the pool is still faulty even after a reboot, how can i fix it?
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
You usually do a pool export/import to fix such problems

If your hardware is not hotplug capable, do a pool export, reboot, pool import
optionally clear error (napp-it menu pools)

If you want an overview of all disks and slots without napp-it, I would
- power down
- remove all disks and power on
- insert a disk, and write down the slot + serial + WWN and remove (on hotplug capable hardware)
- repeat with all disks


With napp-it, just open menu "Disks - SAS2 extension" where all disks are listed with
controller + enclosure + slot number and all other infos like capacity, pool/vdev, Product, WWN, serial and iostat errors.

Print out this page and place it on top of your storage.
On napp-it free you can use an evalkey.
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
Yeah, i took a snapshot of the SAS2 extension page with the slot numbers, but i don't know where are located each slot on the case's backplanes. (There is an identify option that blink a red light but it seems my backplanes are not compatible)

About hotplug, isn't all sata already hotplug capable? If not how can i check this?
(config is X9SRL-F + 3*M1015 + xcase RM424pro + 24*5K4000 hdds)

I tried importing the pool but i got this message:
N7FEus.png
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
Yeah, i took a snapshot of the SAS2 extension page with the slot numbers, but i don't know where are located each slot on the case's backplanes. (There is an identify option that blink a red light but it seems my backplanes are not compatible)

About hotplug, isn't all sata already hotplug capable? If not how can i check this?
(config is X9SRL-F + 3*M1015 + xcase RM424pro + 24*5K4000 hdds)

- Reboot and check if all disks are available and online
- Open menu Pool - import and check if the pool is listed as importable

about hotplug capable
This is not only a electrical problem but the disk controller + driver must support this as well. If you use IBM 1015, they are hotplug capable.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
I found the pool but when i try to import it i get the message in red.

The message complains about a missing device (ie a disk).
This message should only appear if you build a pool from a basic disk that is missing, or
maybe a missing log device.

Have you checked if all disks are really available/ online?
 

Mastaba

Limp Gawd
Joined
Apr 2, 2011
Messages
228
Mhhh, the pool is back online, don't know very well why.
Also i replaced the "failed" disk with a cold spare.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
You can call it Harakiri if you hot unplug one and onother disk from a hardware raid.
With ZFS the pool is mostly online again if enough disks come back.
 

_Gea

2[H]4U
Joined
Dec 5, 2010
Messages
4,025
I have simplified TLS email Setup in OmniOS/OI (needed for alerts over Googlemail)
with napp-it 0.9f1 (May, 30)



#######
step 1: install net-ssleay
#######

Under OI 151a5+, net-ssleay is in the default repository
#################
pkg install net-ssleay

Under OmniOS, net-ssleay is in the perl repository
############
- check publisher: pkg publisher
- if perl.omniti.com is missing:
pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com
pkg install net-ssleay


######
step2: Setup TLS from CPAN (source of Perl modules)
######

perl -MCPAN -e shell
-> and answer yes when it asks if it should try to configure as much as possible by itself

at the cpan prompt, enter:
install Net::SMTP::TLS

-> and answer all questions with the default answer.
install CPAN
reload cpan
exit



#########################################################################
There is a bug with TLS.pm
In napp-it 0.9f1 and up, a modified TLS.pm is included that fix the problem automatically

It changes "SSLv3 TLSv1" to "SSLv3" [remove the TLSv1] in
/usr/perl5/site_perl/t.16.1/Net/SMTP/TLS.pm on line 182

more: https://rt.cpan.org/Public/Bug/Display.html?id=77401
on napp-it < 0.9f1, you must edit TLS.pm manually
#########################################################################
 

spazoid

Limp Gawd
Joined
Jun 16, 2008
Messages
289
Just tried this, and it works without any issues. Omnios 151006 and latest napp-it (may 30th)
 

HammerSandwich

[H]ard|Gawd
Joined
Nov 18, 2004
Messages
1,125
I DLed the current pre-built VM last night. v0.9e1_preview from Jan 12.

Comstar/Targets/Create seems to have the correct link (...&l1=09_comstar%20iscsi&l2=03_targets&l3=01_create%20iscsi%20target), but the page content is the same as Create Target Group. I cannot create targets via Napp-it.
 
Top