Best Practice for configuring ZS3 storage

Hi Team
Can someone help me with the best way to configure data profile for a ZS3-2 storage which is mainly going to be used for oracle database and EBU application.
If we are planning to install ldoms os on storage lun then what data profile should be the best one.
Regards
AB

Without knowing more I would do the following, being very conservative;
Create too pools, with pool0 on controller0 and pool1 on controller1. This is to take advantage of both controllers DRAM, for better performance. Regardless of RAIDZ or RAID 1+0, I would do this.
I am assuming both trays are the same HHDs ( 900G?)
Make the first pool RAID 1+0, with 2 write SSDs and one READ SSD. I would use this pool for redo logs, and database files. Mounting via OISP or dNFS.
Make the second pool RAIDZ, with 2 write SSDs and one READ SSD, I would place my RMAN mount here, and any binary mounts and archive logs. Maybe DB files, if using partitioning. HCC is a great technology to use if you application will benefit from it.
Now for the relaity check!
9 times out of 10 method...
Realistic... just use RAIDZ on both pools. In reality RAID 1+0 generates more IOPS in the array and leads to more DB I/O wait time! It's true!
If you are headed to ECO I'll sit down and explain. short version, RAIDZ does less writes, which means better database performance.
Erik

Similar Messages

Best Practices for Configuration Manager

What all links/ documents are available that summarize the best practices for Configuration Manager?
Applications and Packages
Software Updates
Operating System Deployment
Hardware/Software Inventory

Hi,
I think this may help you
system center 2012 configuration manager best practices
SCCM 2012 task-sequence best practices
SCCM 2012 best practices for deploying application
Configuration Manager 2012 Implementation and Administration
Regards, Ibrahim Hamdy

Best practice for configuring virtual drive on UCS

Hi,
I have two C210 M2 Server with LSI 6G MegaRAID 9261-8i Card with 10 hard drives of each 135GB. When I tried the automatic selection for the RAID configuration, the system has created one virtual drive with RAID 6. My concern is what the best practice to configure virtual drive? Is it RAID 1 and RAID5 or all in one drive with RAID6? Any help will be appreciated.
Thanks.

Since you are planning to have UC apps on the server, Voice applications have specified their recommendations here.
http://docwiki.cisco.com/wiki/Tested_Reference_Configurations_%28TRC%29
I believe your C210 server specs could be matching TRC #1, where you need to have
RAID 1 - First two HDDs for VMware
RAID 5 - Remaining 8 HDDs as datastore for virtual machines ( CUCM and CUC )
HTH
Padma

SAP HCM Implementation: Best Practice for configuring

Hi,
This is my first independent project of HCM implementation. I have just started the system configuration. Done with setting up the PA, PSA, EG and ESG. Assigned to CC. At this stage, I have a very basic question which is, what is the best practice for the next steps of configuration. What do I go to next, step up the OM in the SAP EasyAccess Menu? How should I go from here?
Would really appreciate some explanatory assistance.
Thanks in advance.
Papri
Edited by: papri_rc on Jul 8, 2011 6:40 AM

Its all depends on business requirement
at starting as i advised you review your BBP , make sure you configure everything
for your reference iam giving the following data for OM and PA config..
as part of OM
1.     depict client org structure using simple maintenance , with this you can create large structures in less time(while doing org structure be careful and refer BBP)
2.     maintain integration switches
3.     maintain plan version
4.     maintain number ranges
Configuration for PA
HR Enterprise /PersonnelStructure
u2022     Personnel Areas
u2022     Personnel Sub Areas
u2022     Employee Group
u2022     Employee Sub Group
u2022     Assignment of Personnel Area to Company Code
u2022     Assignment of Employee Sub Group to Employee Group
Basic Settings
1.     Maintain Number Range Intervals for Personnel Numbers
2.     Determine defaults for number ranges
Personal Data
1.     Create Forms of Address
2.     Create Marital Status
Family
1.     Defined Possible Family Members
Addresses
1.     Create Address Type
Communication
1.     Create Communication Types
Contractual and Corporate Agreements
1.     Define Contract Types
2.     Determine periods of notice
Employee Qualifications
1.     Create education establishment types
2.     Define Education Training
3.     Create educational Certificates
4.     Create branches of study
5.     Determine permissible certificates for education type
Infotype Menus
1.     User Group Dependency on Menus and Info groups
2.     Infotype Menu
3.     Determine choice of Infotype menus
4.     Infotype Menus
Actions
1.     User Group Dependency on Menus and Infogroups
2.     Info Group
3.     Personnel Action Types
4.     Create reasons for personnel actions
5.     Change Action Menu
*Developments(ABAPconsultant will do)     *Field Enhancements     (any field enhancements in infotypes)
Customer Infotypes     -Develop any customer infotypes if required for the business from 9000 series
Edited by: Piscian . on Jul 8, 2011 9:08 AM

Best practice for presenting different storage vendors arrays..?

Hi, can't seem to find a specific answer in writing on this.
Typical FlexPod already deployed, with different vsans on different n5ks.
Uplinks from UCS / NetApp are FCoE.
So I now want to present some normal FC (not FCoE, nexsan) storage to the same blades on the UCS system via the n5ks.
My main questions are can I use same vsans / vHBAs as for NetApp storage, or should you be creating different vsans and vHBAs for different storage vendors?
If multiple vsan is recommended, is Trunking multiple vsans over FCoE from UCS to FC fine, I'm sure it is but just want some experience on it.
Thanks!

Thanks guys for the input, interesting to see there is a difference of opinion on this topic. Makes me feel happier that I'm not being too stupid and someones posted back a link to a best practice guide on the subject that I've blatantly missed! (still time for that no doubt!)
Like I've said, I haven't found anything specific in writing on this, but keeping vendors on seperate vsans / hbas feels like the best way to do it anyway with only a little extra config in the grand scheme of things.

Best practices for configure Rogue Detector AP and trunk port?

I'm using a 2504 controller. I dont have WCS.
My questions are about the best way to configure a Rogue Detector AP.
In my lab environment I setup the WLC with 2 APs. One AP was in local mode, and I put the other in Rogue Detector mode.
The Rogue Detector AP was connected to a trunk port on my switch. But the AP needed to get its IP address from the DHCP server running on the WLC. So I set the native vlan of the trunk port to be the vlan on which the WLC management interface resides. If the trunk port was not configured with a native vlan, the AP couldn't get an address through DHCP, nor could the AP communicate with the WLC. This makes sense because untagged traffic on the trunk port will be delivered to the native vlan. So I take it that the AP doesn't know how to tag frames.
Everything looked like it was working ok.
So I connected an autonomous AP (to be used as the rogue), and associated a wireless client to it. Sure enough it showed up on the WLC as a rogue AP, but it didn't say that it was connected on the wire. From the rogue client I was able to successfully ping the management interface of the WLC.
But the WLC never actually reported the rogue AP as being connected to the wired network.
So my questions are:
1. What is the correct configuration for the trunk port? Should it not be configured with a native vlan? If not, then I'm assuming the rogue detector AP will have to have a static IP address defined, and it would have to be told which vlan it's supposed to use to communicate with the WLC.
2. Assuming there is a rogue client associated with the rogue AP, how long should it reasonably take before it is determined that the rogue AP is connected to the wired network? I know this depends on if the rogue client is actually generating traffic, but in my lab environment I had the rogue client pinging the management interface of the WLC and still wasn't being picked up as an on-the-wire rogue.
Thanks for any input!!

#what's the autonomous AP's(as Rogue AP) Wired and Wireless MAC address?
it has to be +1 or -1 difference. If Wired MAC is x.x.x.x.x.05 and the wireless mac should be x.x.x.x.x.04 or 06. It is not going to detect if the difference is more than + 1 or - 1.
#Does the switch sees the Rogue AP's wired MAC on its MAC table.
Rogue Detector listens to ARPs to get all the Wired MAC info and forwards to WLC, It compares with Wireless MAC, if there is a +1 or -1 difference then it will be flagged as Rogue on wire. And the client that connected to it is also marked as found on wire.
Regards to Trunking, Only Native vlan matters per trunk link, just configure the right vlan as native and we're done.
It is not mandatory to keep the Rogue detector on Management vlan of wlc. It can also be on L3 vlan also as long as it can join the WLC to forward the learnt wired MACs.
So if we don't have +1, -1 difference on Rogues then you've to use RLDP which will work with your existing setup to find Rogue on wire. there's a performance hit when we use this feature on local mode APs.
Note: For AP join - AP can't understand Trunk, meaning if AP connected to Trunk it'll only talk to its native vlan irrespective of AP mode, however rogue detector listens to the Trunk port to learn MACs via ARPs from different VLANs and forwards to WLC using native vlan.

Best Practices for configuring ICMP from the outside

Question,
Are there any best practices or best recommendations on how ICMP should be configured from the outside? I have been cleaning up the rules on our ASA as a lot were simply ported over years ago when we retired our PIX. I noticed that there is a rule to allow ICMP any any and began to wonder how this works when the rules above are specific IP addresses and specific ports. This in thurn started me looking to see if there was any documentation or anything to help me determine a best practice. Anyone know of anything?
As a second part how does this flow on a firewall if all the addresses are natted? It the ICMP traffic simply passed through the NAT and the destiantion simply responds?
Brent

Here you go, bro!
http://checkthenetwork.com/networksecurity%20Cisco%20ASA%20Firewall%20Best%20Practices%20for%20Firewall%20Deployment%201.asp#_Toc218778855
access-list inside permit icmp any any echo
access-list inside permit icmp any any echo-reply
access-list inside permit icmp any any unreachable
access-list inside permit icmp any any time-exceeded
access-list inside permit icmp any any packets-too-big
access-list inside permit udp any any eq 33434 33464
access-list deny icmp any any log
P/S: if you think this comment is useful, please do rate them nicely :-)

Best Practice for VOD Remote Storage

Hi All,
We have a couple of servers that host FLV videos for on-demand playback. The file storage however is on a separate server, and all the stored VOD content is accessed via UNC share, ie \\servername\share\foo\bar. The servers have a fast connection to the file server, however, whenever we try to play videos from the FMS servers, the first time its requested, they play back jerky and jumpy. Most of the time though, the second time the video is played, it plays fine. Also, if you try to seek ahead to a spot where no one's played back from before, it will be jerky the first time, but fine the second time.
I'm assuming that the first time through, the FMS server caches the content, and then the 2nd time it servers it from the cache. Since the servers have a very fast connection between them, what can be done to solve this?
Also, would it be better to have an Origin server host the file content locally, and then use Edge servers to connect to and serve the files? Would this alleviate the problem at all? Thank you for any help.
Best,
-Costin

We're probably going to have to define 'fast connection' a bit here. It sounds like either the speed of requests needed for FMS, or their size is causing an issue getting sufficient data flow to FMS to avoid stuttering on the desired file. There's a few ways that you can measure this. You can go the route of measuring the I/O on network from your UNC server location, or if you're running an FMIS file adaptor you can plug in and log the speed of return of your reads. At the end of it though it's going to be about the numbers of how quickly we're able to answer FMS' requests for data and demonstrating that we do or don't have enough flowing through here.
Asa

Best Practices for Configuring Windows Server 2012 R2 for VM Performance

Good afternoon,
I've been told that VMWare has written a document or article on how to get the most performance out of an instance of Server 2012 R2 as a virtual machine. I was told this magical article has things like a checklist of features or services that can be turned off (such as tablet features) that will increase the performance/low overhead on the VM.
Can anyone help me find such an article?
Thanks,
Johnathan

Hello there,
There is no upgrade path from Server 2012 Essentials to R2. R2 is not an upgrade, it is a new build. Essentials does not upgrade easily, you will want to perform a migration of the server. This will require additional hardware. You
can possibly backup the 2012 Essentials, build a temp server on workstation hardware just to do the migration from, restore the backup to the new server and then perform the migration to R2.

Networking "best practice" for setting up a farm

Hi all.
We would like to set an OracleVM farm, and I have a question about "best practice" for
configuring the network. Some background:
- The hardware I have is comprised of machines with 4 gig-eth NICs each.
- The storage will be coming primarily from a backend NAS appliance (Netapp, FWIW).
- We have already allocated a separate VLAN for management.
- We would like to have HA capable VMs using OCFS2 (on top of NFS.)
I'm trying to decide between 2 possible configurations. The first would keep physical separation
between the mgt/storage networks and the DomU networks. The second would just trunk
everything together across all 4 NICs, something like:
Config 1:
- eth0 - management/cluster-interconnect
- eth1 - storage
- eth2/eth3 => bond0 - 8021q trunked, bonded interfaces for DomUs
Config 2:
- eth0/1/2/3 => bond0
Do people have experience or recommendation about the best configuration?
I'm attracted to the first option (perhaps naively) because CI/storage would benefit
from dedicated bandwidth and this configuration might also be more secure.
Regards,
Robert.

user1070509 wrote:
Option #4 (802.3ad) looks promising, but I don't know if this can be made to work across
separate switches.It can, if your switches support cross-switch trunking. Essentially, 802.3ad (also known as LACP or EtherChannel on Cisco devices) requires your switch to be properly configured to allow trunking across the interfaces used for the bond. I know that the high-end Cisco and Juniper switches do support LACP across multiple switches. In the Cisco world, this is called MEC (Multichassis EtherChannel).
If you're using low-end commodity-grade gear, you'll probably need to use active/passive bonds if you want to span switches. Alternatively, you could use one of the balance algorithms for some bandwitch increase. You'd have to run your own testing to determine which algorithm is best suited for your workload.
The Linux Foundation's Net:Bonding article has some great information on bonding in general, particularly on the various bonding methods for high availability:
http://www.linuxfoundation.org/en/Net:Bonding

Best practice for running multiple sites on 1 CF install?

Hi-
I'm setting up a new hosting environment (Windows Server 2008 Standard 64 bit VPS configuration, MySQL, IIS 7, CF 9)
Has anyone seen any docs or can anyone suggest best practices for configuring multiple sites in this environment? At this point I'm thinking simple is best, one new site in IIS for each client (domain) and point it to CF.
Given this environment, is anyone aware of any gotchas within the setup of CF 9 on IIS 7?
Thank you in advance,
Rich

There's nothing wrong with that approach. You can run as many IIS sites as you like against a single CF install.
As for installing CF on IIS 7, I recommend that you do the following: install CF 9 without connecting it to IIS, then installing the 9.0.1 upgrade and any hotfixes, then connecting CF to IIS using the web server configuration utility. This will keep you from having to install the IIS 6 compatibility layer that's needed with CF 9 but not with CF 9.0.1.
Dave Watts, CTO, Fig Leaf Software
http://www.figleaf.com/
http://training.figleaf.com/

Best practices for implementing OIM

We plan on putting OIM servers behind LB (hardware). When I develop OIM client I am required to specify OIM endpoint(s) via property java.naming.provider.url. In case of LB I'd specify a virtual host there. The question is what is the best practice for configuring LB - timeout, persistence, monitoring? I don think LB vendor is relevant, but just in case, I have a choice of F5 BigIP and Citrix Netscaler.
My understanding is that Java class tcUtilityFactory is supposed to be instantiated once (in a web client) and maintain the connection, but LB will close the connection after timeout is exceeded. So another question is if I want to use LB I have to take care of rebuilding connection when it is expired, or open/close connection every time tcUtilityFactory is needed. Any advice will be appreciated.
Thanks,
Alex

No i was not going to sync timeouts - just let it close connections after say, 5 min of inactivity. The reason is that performance data is horrible - from my desktop environment, initialization takes almost 9 sec, while reading data from OIM - only 150 milliseconds. I can't afford more than .5 sec on the whole OIM operation, as we are talking about customer experience.
Thanks,
Alex

Best practice for server configuration for iTunes U

Hello all, I'm completely new to iTunes U, never heard of this until now and we have zero documentation on how to set it up. I was given the task to look at best practice for setting up the server for iTunes U, and I need your help.
*My first question*: Can anyone explains to me how iTunes U works in general? My brief understanding is that you design/setup a welcome page for your school with sub categories like programs/courses, and within that you have things like lecture audio/video files and students can download/view them on iTunes. So where are these files hosted? Is it on your own server or is it on Apple's server? Where & how do you manage the content?
*2nd question:* We have two Xserve(s) sitting in our server room ready to roll, my question is what is the best method to configure them so it meets our need of "high availability in active/active mode, load balancing, and server scaling". Originally I was thinking about using a 3rd party load balancing device to meet these needs, but I was told there is no budget for it so this is not going to happen. I know there is IP Failover but one server has to sit in standby mode which is a waste. So the most likely scenario is to setup DNS round robin and put both xserves in active/active. My question now is (this maybe related to question 1), say that all the content data like audio/video files are stored by us, (We are going to link a portion of our SAN space to Xserve for storage), if we are going with DNS round robin and put the 2 servers in Active/Active mode, can both servers access a common shared network space? or is this not possible and each server must have its own storage space? And therefore I must use something like RSYNC to make sure contents on both servers are identical? Should I use XSAN or is RSYNC good enough?
Since I have no experience with iTunes U whatsoever, I hope you understand my questions, any advice and suggestion are most welcome, thanks!

Raja Kondar wrote:
wht is the Best Practice for having server pool i.e
1) having a single large serverpool consisting of "n" number of guest vm
2) having a multiple small serverpool consisting of less of number of guest vm I prefer option 1, as this gives me the greatest amount of resources available. I don't have to worry about resources in smaller pools. It also means there are more resources across the pool for HA purposes. Not sure if this is Official Best Practice, but it is a simpler configuration.
Keep in mind that a server pool should probably have up to 20 servers in it: OCFS2 starts to strain after that.

Best Practice for SSL in Apache/WL6.0SP1 configuration?

What is the best practice for eanbling SSL in an Apache/WL6.0SP1
configuration?
Is it:
Browser to Apache: HTTPS
Apache to WL: HTTP
or
Browser to Apache: HTTPS
Apache to WL: HTTPS
The first approach seems more efficient (assuming that Apache and WL are
both in a secure datacenter), but in that case, how does WL know that the
browser requested HTTPS to begin with?
Thanks
Alain

A getScheme should return HTTPS if the client is using HTTPS or HTTP if it
is using HTTP.
The option for the plug-in to use HTTP or HTTPS when connecting to Weblogic
is up to you but regardless the scheme of the client will be passed to
WebLogic.
Eric
"Alain" <[email protected]> wrote in message
news:[email protected]..
How should we have the plug-in tell wls the client is using https?
Should we have the plugin talk to wls in HTTP or HTTPS?
Thanks
Alain
"Jong Lee" <[email protected]> wrote in message
news:3b673bab$[email protected]..
The apache plugin tells wls the client is using https and also pass on
the
client
cert if any.
"Alain" <[email protected]> wrote:
What is the best practice for eanbling SSL in an Apache/WL6.0SP1
configuration?
Is it:
Browser to Apache: HTTPS
Apache to WL: HTTP
or
Browser to Apache: HTTPS
Apache to WL: HTTPS
The first approach seems more efficient (assuming that Apache and WL
are
both in a secure datacenter), but in that case, how does WL know that
the
browser requested HTTPS to begin with?
Thanks
Alain

Best practices for dealing with Exceptions on storage members

We recently encountered an issue where one of our DistributedCaches was terminating itself and restarting due to an RuntimeException being thrown from our code (see below). As usual, the issue was in our own code and we have updated it to not throw a RuntimeException under any circumstances.
I would like to know if there are any best practices for Exception handling, other than catching Exceptions and logging them. Should we always trap Exceptions and ensure that they do not bubble back up to code that is running from the Coherence jar? Is there a way to configure Coherence so that our DistributedCaches do not terminate even when custom Filters and such throw RuntimeExceptions?
thanks, Aidan
Exception below:
2010-02-09 12:40:39.222/88477.977 Oracle Coherence GE 3.4.2/411 <Error> (thread=DistributedCache:StyleCache, member=48): An exception (java.lang.RuntimeException) occurred reading Message AggregateFilterRequest Type=31 for Service=DistributedCache{Name=StyleCache, State=(SERVICE_STARTED), LocalStorage=enabled, PartitionCount=1021, BackupCount=1, AssignedPartitions=201, BackupPartitions=204}
2010-02-09 12:40:39.222/88477.977 Oracle Coherence GE 3.4.2/411 <Error> (thread=DistributedCache:StyleCache, member=48): Terminating DistributedCache due to unhandled exception: java.lang.RuntimeException

Bob - Here is the full stacktrace:
2010-02-09 13:04:22.653/90182.274 Oracle Coherence GE 3.4.2/411 <Error> (thread=DistributedCache:StyleCache, member=47): An exception (java.lang.RuntimeException) occurred reading Message AggregateFilterRequest Type=31 for Service=DistributedCache{Name=StyleCache, State=(SERVICE_STARTED), LocalStorage=enabled, PartitionCount=1021, BackupCount=1, AssignedPartitions=205, BackupPartitions=204}
2010-02-09 13:04:22.653/90182.274 Oracle Coherence GE 3.4.2/411 <Error> (thread=DistributedCache:StyleCache, member=47): Terminating DistributedCache due to unhandled exception: java.lang.RuntimeException
2010-02-09 13:04:22.653/90182.274 Oracle Coherence GE 3.4.2/411 <Error> (thread=DistributedCache:StyleCache, member=47):
java.lang.RuntimeException: java.lang.ClassNotFoundException: com.edmunds.vehicle.Style$PublicationState
     at com.edmunds.common.coherence.EdmundsEqualsFilter.readExternal(EdmundsEqualsFilter.java:84)
     at com.tangosol.io.pof.PortableObjectSerializer.initialize(PortableObjectSerializer.java:153)
     at com.tangosol.io.pof.PortableObjectSerializer.deserialize(PortableObjectSerializer.java:128)
     at com.tangosol.io.pof.PofBufferReader.readAsObject(PofBufferReader.java:3284)
     at com.tangosol.io.pof.PofBufferReader.readAsObjectArray(PofBufferReader.java:3328)
     at com.tangosol.io.pof.PofBufferReader.readObjectArray(PofBufferReader.java:2168)
     at com.tangosol.util.filter.ArrayFilter.readExternal(ArrayFilter.java:243)
     at com.tangosol.io.pof.PortableObjectSerializer.initialize(PortableObjectSerializer.java:153)
     at com.tangosol.io.pof.PortableObjectSerializer.deserialize(PortableObjectSerializer.java:128)
     at com.tangosol.io.pof.PofBufferReader.readAsObject(PofBufferReader.java:3284)
     at com.tangosol.io.pof.PofBufferReader.readAsObjectArray(PofBufferReader.java:3328)
     at com.tangosol.io.pof.PofBufferReader.readObjectArray(PofBufferReader.java:2168)
     at com.tangosol.util.filter.ArrayFilter.readExternal(ArrayFilter.java:243)
     at com.tangosol.io.pof.PortableObjectSerializer.initialize(PortableObjectSerializer.java:153)
     at com.tangosol.io.pof.PortableObjectSerializer.deserialize(PortableObjectSerializer.java:128)
     at com.tangosol.io.pof.PofBufferReader.readAsObject(PofBufferReader.java:3284)
     at com.tangosol.io.pof.PofBufferReader.readObject(PofBufferReader.java:2599)
     at com.tangosol.io.pof.ConfigurablePofContext.deserialize(ConfigurablePofContext.java:348)
     at com.tangosol.coherence.component.util.daemon.queueProcessor.Service.readObject(Service.CDB:4)
     at com.tangosol.coherence.component.net.Message.readObject(Message.CDB:1)
     at com.tangosol.coherence.component.net.message.requestMessage.distributedCacheRequest.partialRequest.FilterRequest.read(FilterRequest.CDB:8)
     at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache$AggregateFilterRequest.read(DistributedCache.CDB:4)
     at com.tangosol.coherence.component.util.daemon.queueProcessor.service.Grid.onNotify(Grid.CDB:117)
     at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache.onNotify(DistributedCache.CDB:3)
     at com.tangosol.coherence.component.util.Daemon.run(Daemon.CDB:37)
     at java.lang.Thread.run(Thread.java:619)
Caused by: java.lang.ClassNotFoundException: com.edmunds.vehicle.Style$PublicationState
     at java.lang.Class.forName0(Native Method)
     at java.lang.Class.forName(Class.java:169)
     at com.edmunds.common.coherence.EdmundsEqualsFilter.readExternal(EdmundsEqualsFilter.java:82)
     ... 25 more
2010-02-09 13:04:23.122/90182.743 Oracle Coherence GE 3.4.2/411 <Info> (thread=Main Thread, member=47): Restarting Service: StyleCacheOur code was doing something simple like
catch(Exception e){
    throw new RuntimeException(e);
}Would using the ensureRuntimeException call do anything for us here?
Edited by: aidanol on Feb 12, 2010 11:41 AM

Best Practice for configuring ZS3 storage

Similar Messages

Maybe you are looking for