Next Version? Topic is solved

FAQ, getting help, user experience about FancyCache
timerrington
Level 1
Level 1
Posts: 1
Joined: Thu Jun 27, 2013 4:15 am

Re: Next Version?

Post by timerrington »

HI Support

The v0.90 attachings in this thread - Is this the Disk Version or Volume Verson?

I can't seem to start as many caches with this version as the 0.80 disk version.

Cheers

Tim Errington
twu2
Level 2
Level 2
Posts: 7
Joined: Wed Jun 26, 2013 4:12 am

Re: Next Version?

Post by twu2 »

support wrote:@twu2,

We have checked the dump file and found a very strange error that caused the bluescreen. May I have your disk/volume information, memory installed, and cache configurations?

According to the dump file, this strange error happened with Defer-Write. Could you try cache without Defer-Write enabled?
2 disks, 32GB RAM. the cache configurations like the attachment (only the cache size is 15GB).
I don't test it without defer-write, but I can use it when change the cache size to 14400MB.
Attachments
primocache.png
primocache.png (110.52 KiB) Viewed 65174 times
minhgi
Level 10
Level 10
Posts: 255
Joined: Tue May 17, 2011 3:52 pm

Re: Next Version?

Post by minhgi »

support wrote:@minhgi,
Thank you very much! You're right. There is an overflow that will cause the crash if the partition to be cached is too big. I'm sorry that I made a mistake on the supported maximum partition size. PrimoCache aslo forgot to do a full parameter check before starting cache. This issue will be fixed in the next version. Here is the list of supported maximum partition size.

Cache Block Size Maximum Partition Size
4KB 4TB
8KB 8TB
.... ...
256KB 256TB

@turkina,
Do multiple partitions share one cache in your scenario? And each partition is less than 4TB? If there's a partition larger than 4TB, it is better to change the block size to a larger value.
@ Support

Here is my final Primocache configuration of my workstation. So far it have been very stable for the past 24 hours.

Hardware -
CPU: Intel I7-3770K
MB: ASUS Maximus V Formula
Memory: OCZ 16GB 2133 9-9-9-9
Harddrive: LSI 9260-8i Fastpath & Cachcade 2.0 / OCZ Vector 2x128GB Raid 0 /Seagate Savvio 15k.2 2x73GB Raid 1 / Seagate 4x2TB Raid 5 / defer write 4 second on Raid controller
Primocahe: L1 2GB / L2 128GB using OCZ Vector and fastpath / 4.02 GB Overhead / 8K Cluster Size / Defer write 4 seconds

Caution - if the computer is using a hard raid controller with defer write and planning on using the defer write in Primocache, set the defer write, in Primocache, equal to or less than the one on the controller. The computer will hang or crash on shutdown. The raid controller will try to flush data earlier to hard drive and sent a shut down command to the computer. Primocache had not the chance to flush its data. The result is computer hang or corrupt data.

Suggestion to make Primocache better -
1) Allow option for hotdata , spill over, or file/folder specific caching algorithm (I prefer the spill over cache that is currently in version 0.90)
2) Allow option to select different caching algorithm for the L2 cache like version 0.8
3) Allow option to flush L1 cache to L2 cache immediately
4) Allow option to set average write time to L2 cache.

This is all I have for now. I will update once I have any other idea or bug found. So far, Primocache had meet my expectation and looking forward to a complete product. :D
Attachments
2013-06-27_064912.png
2013-06-27_064912.png (64.2 KiB) Viewed 65161 times
Last edited by minhgi on Fri Jun 28, 2013 5:03 am, edited 1 time in total.
User avatar
Support
Support Team
Support Team
Posts: 3623
Joined: Sun Dec 21, 2008 2:42 am

Re: Next Version?

Post by Support »

@twu2,

When you set cache size to 15GB, did system crash immediately? Had you ever clicked "Pause/Resume" button?
twu2 wrote:2 disks, 32GB RAM. the cache configurations like the attachment (only the cache size is 15GB).
I don't test it without defer-write, but I can use it when change the cache size to 14400MB.
twu2
Level 2
Level 2
Posts: 7
Joined: Wed Jun 26, 2013 4:12 am

Re: Next Version?

Post by twu2 »

support wrote:@twu2,
When you set cache size to 15GB, did system crash immediately? Had you ever clicked "Pause/Resume" button?
yes, the os crash immediately.

I made some test tonight.

1. cache size 15360MB without defer-write, it will work.
2. cache size 14400MB with defer-write, it will work.
3. cache size 15360MB with defer-write, got bluescreen after click start button.
Incriminated
Level 6
Level 6
Posts: 65
Joined: Fri May 31, 2013 3:03 pm

Re: Next Version?

Post by Incriminated »

Primocahe: L1 2048GB <--- cool DDR8, get back to the future :D

1. you really have a hardware-raid or FW-raid chip on your motherboard that just "look" like a hardware raid for a simple mind? ah ok i see it is a hardware controller indeed, sry - but "good", why you want a raid and cache? any realistic use case that needs that hardly? <-- these hint is for the others, chances with software or mobo-raid are nearly zeroed.
2. what do you mean with defer-write of the raid? a raid doesn't defer writes externally as far as i know, everything happens internally, it reports a usual sata-drive externally, but raids are not supported in fance/primo-cache, so that might me the simple cause or the HW-raid FW having some incompatibilities in properly reporting a usual normal drive externally or interpreting usual normal drive commands externally.
(( 3. What do you mean with a "shut" command. "shut down"? or "shut up"? ;) only a joke ))
The raid controller will try to flush data earlier to hard drive and sent a shut command to the computer.
(( Caution: I have a poo without having anything eaten .... :D only a joke ))
(( http://www.youtube.com/watch?v=HQoRXhS7vlU :o just to get you in the right mood ;) ))

the answer is not a mystery to see here:
Image
In your case to BIOS and OS the truely HW-raid controller should report as a disk without a special disk driver. (C:\Windows\system32\driver\disk.sys)

How is the Raid-controller able to want to flush the data to the disk before primocache passed it on to the controller!?
Anything you write in the OS is intercepted by primocache for defer-write... nothing reaches the controller/disks before that. FIrstly when primocache flushes cache that data is passed on to the raid. So what you say cannot be true! the raid is unable to try to defer anthing to the disks it doesnt even know of. Think about it again! YOu write everything to primocache... no more the raid, and when primocache flushes then firstly the raid is accessed for write.

The more likely is that a systray or service application or simply the special raid-driver of your raid-controller tries to send internal commands... but that drive is "primocache" on top what puts any internal-raid command at high risk, and by that additionally anything the raid-controller sends without a reason to emulate a standard drive to the os, cannot be properly interpreted by the OS because it only listens to "primocache" for that drive... primocache cannot listen or send to propriety third party layers, what the raids-controller sends is invalid for a direct write - it communicates through primocache!!! A non-standard management command lands in the cache for at least 10 seconds, while that management app trys to read that "result" immediately after - which demand is passed on immediately to the raid (1st read) Tadaaa, and thats why romex doesnt support these, because primocache isnt able to provide a layer for any raid out there to directly pass on or receive internal raid-driver that "make up" a standard-drive :roll:

A raid-controller must be standalone, by that means flawlessly work with your individual sata-controller driver like any other drive without any special necccessities, in special software-layers, that is the only way a software like "primocache" might work....might, because the programm actually is without any warrant in general... not saying it not working wonderful with many standard-drive. no issue here, less known.

Sad to see such great hardware in hands of - sry to say it - newbies (no offense) that warn people with half knowledge in bold-letters :geek: .

*// Just found these according to the LSI shut downs with cache-solutions:
http://alturl.com/nuzrk - SaxGuy explains it very well *// just for relaxing a bit, bro, you are demanding impossible actually.



PS Im not an employee
Last edited by Incriminated on Fri Jun 28, 2013 4:34 am, edited 35 times in total.
User avatar
intika
Level 9
Level 9
Posts: 178
Joined: Wed Oct 13, 2010 1:10 am

Re: Next Version?

Post by intika »

Verry good news :) just downloaded, and start testing .
thanks :)
User avatar
intika
Level 9
Level 9
Posts: 178
Joined: Wed Oct 13, 2010 1:10 am

Re: Next Version?

Post by intika »

Nice version, look just have some cosmetic changes lool (i am kidding)
Why Invisible memory is not enabled ?
minhgi
Level 10
Level 10
Posts: 255
Joined: Tue May 17, 2011 3:52 pm

Re: Next Version?

Post by minhgi »

@ Incriminated
Sad to see such great hardware in hands of - sry to say it - newbies (no offense) that warn people with half knowledge in bold-letters :geek: .
I see the error in my words and corrected the "2048GB / shut up command" grammar mistake. I am sure that someone out there will try to have both hardware/software defer write for fun.

Lol. The point is.... Don't worry about other people toys. You either takes the caution as a warning or move on. No need to take the hard written facts so serious. link link this and that. that show him. Typical young boi behavior. BehaVE NOW. sORRY that I Did not type proper graMMer.
User avatar
Support
Support Team
Support Team
Posts: 3623
Joined: Sun Dec 21, 2008 2:42 am

Re: Next Version?

Post by Support »

@timerrington,

This is a volume edition. No disk edition will be released. You can select multiple partitions with one cache.
timerrington wrote:The v0.90 attachings in this thread - Is this the Disk Version or Volume Verson?
I can't seem to start as many caches with this version as the 0.80 disk version.
Post Reply