 |
How much TB can a volume contain? |
 |
The firmware uses a 64bit block offset. In theory we can create a volume of 2^64 x 512 byte blocks. However, in firmware our volumes are internally limited to 128TB.
This can be increased, at the cost of using more memory resources on the controller. In practical terms our volumes are limited by the number of available disks which can fit into the attached enclosures.
At the moment this is limited to 6 expansion enclosures. This is also an internal limit in the software which can be increased but there is a cost in memory resources and increased system testing permutations. |
 |
How many volumes are in an array? |
|
Single RAID is allowed to have up to 112 disks. There can be up to 64 RAIDS configured on the system and up to 1024 Logical volumes. |
 |
Does your controller support SMART? |
|
No.
Our system turns off SMART on the disks. This is because with SMART enabled the disks sometimes do not respond to read and write commands for up to 15 seconds. This causes problems for video and other applications which cannot tolerate a high command latency. Instead what we do is maintain our counts of disk errors and compare these errors to thresholds. If these thresholds are exceeded we indicate a fault with the disk. |
 |
Does your controller support disk scrubbing and disk cloning? |
|
Yes.
Disk Scrubbing We understand this to mean the background checking of disk media to find latent medium errors on otherwise functional disks. We support this. Our system support automatic scheduled background verification of RAID volumes which checks parity and the disk media. Any media errors or parity errors are found and fixed. If a hard media error is found this block will be reallocated to a different area on the same disk.
Disk Cloning, If an excessive number of faults are found on the a disk, or if the SATA/SAS connection shows too many link errors the disk is deemed to be faulty even though it remains active within a RAID. In this case the system will move the data off the failing disk onto a spare, before the disk fails outright. This, we believe is the same to the "disk cloning" feature offered by other manufacturers. |
|
Does controller support 1GB Ethernet? |
|
- Power off the RAID subsystem and remove the power cords. After 3 minutes, power on the RAID subsystem. Check if it will boot normally.
- Another option is to power off the RAID subsystem. Remove the BBM from the controller module. Check if the controller can boot normally. If the controller boot up is normal, reconnect the BBM and check again.
- To avoid this boot up issue with BBM, upgrade the controller firmware to version 1.45.
|
 |
In dual controller mode, can both controllers write to the same volume? In other words, can both controllers access the same volume? |
|
Yes. |
 |
Do your controllers support MPIO drivers? |
|
Yes, we have an MPIO driver in development. |
 |
Is the SNMP support of your controller full-featured or limited? Does it contain MIB files? |
|
All we support is SNMP traps - ALERT messages encoded and sent on UDP port 162. We don't support anything else on SNMP. |
 |
Does your controller send/receive (during normal operations, not during an event) the status of the enclosure? For example, capacity consumed, how many volumes, volumes mapping, etc? |
|
EP-3169 and EP-4429 series of controllers provide this status whenever asked for it. All these items of information are available via the proRAID Manager Web GUI. |
 |
What is the embedded OS on the controller? |
|
The embedded operating system is QNX Neutrino. |
 |
What is the advantage of using the QNX Neutrino OS? |
|
QNX Neutrino is a genuine real time OS, unlike Linux. It is also much more lightweight than Linux, requiring less resources and hence you get more out of the CPU for your application. Furthermore if there is a problem with the OS we have a support contract with QNX. Getting support for Linux often involves sending out requests on user-supported mailing lists. Even when a support contract exists, the providers of this support need to be able to analyze and fix problems in code which they did not write. QNX Neutrino considerably reduces the complexity of permutations of different kernel & drivers you get with Linux. QNX Neutrino has a reputation for being fast, robust and fault-tolerant. As such it is a good choice for our application. QNX Neutrino is POSIX compliant and our code has been ported using POSIX calls. This means that if we ever need to move to Linux will be easy for us to do so. |
 |
Can your firmware now support multiple JBOD expansion? If yes, how many Jbod can be supported? |
|
Yes. We can support 5 expansion Jbod (6 enclosures in total) for 16 bay model and 1 expansion Jbod for 42bay model (2 enclosures in total). |
 |
Can your firmware now support redundant controllers? |
|
Yes. |
 |
When does rebuild start after a disk is drop off from the system? |
|
The rebuild will starts after 5 minutes, the reasons for this Is, there are tendencies that a user mistakenly remove a drivesoOr a drives is not totally detective it only need a disk reset command. |
