MPIO - slow speed

some fresh news (RN4220 NTGR highest device!!!!):

even with 6.4.0 T34 2*1gbe for Write is unavailable - only ~50-60MB/sec write speed with awesome latency (300-2000), so the storage is almost unusable.

w/o MPIO (1 NIC) it's a little faster and can work at 60-70MB/sec with the same shoking latency, but all other services like SMB or GUI is out of service during iSCSI high load.

conclusion is simple - at these days NTGR devices has no MPIO and very limited iSCSI support. and this is not Enterprise nor MID-Enterprise level devices.

I do not know what does your QA department do, but obviously something wrong.

new case

25526961

RN4220 is inaccessible within default subnet

incredible support team answer:

"if device is not halted now I can't escalate case, let's wait.

are Netgear still joking?

or Russian Support focused on home routers with "reset" option for any case???!!

I understand the case was escalated to L3 support and some recommendations were made regarding the network layout. Please let us know if you are seeing improvements now.

Hello.

Yes, the case was escalated and some RDN network config changes were maid.

at this moment:

I've achived ~110MB/sec sequential read, and ~90MB/sec sequential write with one 1gbe NIC.

let it be the best for RN4220.

I still can't use MPIO at 4220 because of errors during initial connection (but five years old!!! RND6 pro work "well" with MPIO) and now I've lost any iscsi connetion to 4220 - will open new case. or maybe just reboot it several times? sad smile.

I will tell you honestly - I'm tired. every f*cking day I'm reading tons of email alerts from backup software which can't write/read/access 4220.

I'm opeing cases, chatting with L3 support, rebooting, reconfiguring, updating all around 4220.

I have eight (!!!!!) RDN6 boxes and it's 4-5 years old, but it's WORKING. EVERY SINGLE DAY TILL 5 YEARS!

anyway I finally convinced about NTRG enterprise devices quality - will never buy it again.

new bug after reboot

endless "reconnecting to readynas admin page" and blue progress bar.

SMB shares are available, but iSCSI not (even one path).

RAIDar can't find any RDN at the network.

one more case?

OK!!

The existing case is still open, pending feedback from you, so you don't need to open a new case. Please continue to work with the L3 tech under that case.

Hello.

72 hours uptime without issues!

it's new high score.

thanks everyone (especially L3 tech's Justin and Stefano).

That's good news, thanks for the update.

As OptimusPrime mentioned, we are working on getting an MPIO KB article published.

What do you think the solution was in your environment? It may help others who come across this thread.

Hello BrendanM.

There were two misconfigs:

First

subnets for MPIO must be sepatated (i.e. 192.168.0.0/24 and 192.168.1.0/24).

other vendors allows one subnet for MPIO. it's usually separated VLAN for iSCSI traffic only.

Second

gateway at NICs config - GUI doesn't allow to save 0.0.0.0 at GW string (and that is correct!), but at the same time it's sets it (0.0.0.0) automatically.

it's weird and confusing. no GW is NO GW, nor 0.0.0.0 nor something else.

besause no one in sane mind will not route iSCSI traffic across routers.

other wishes:

- OOBM or other dedicated management

- VLANs