#osn 1800

[thunderlinkcom]

Why throughput is low when using low frame sizes on a OSN1800V

Issue Description

Customer has noticed some issues on their Ethernet composite services configured on a OSN1800. Their service was a 10G interface (VPLS) with multiple 1G interfaces (PWE3) at the remote NE`s.

He reported low throughput when using low frame sizes (Eg 64bytes).

His main question was what possible causes can make this odd behavior. 

Solution

The following is a standard Layer 1 802.1Q frame where we can see the 20 bytes header(IFG, Preamble and SFD) included.

The same packet if it is in Layer 2, the Layer 1 header is not included, and it looks like this:

So the first concept is Layer 1, Layer 2, Layer 3 packet length. The Layer 1 packet length is the whole packet length included in the Layer 1 header, the Layer 2 packet length will exclude the Layer 1 header and the Layer 3 packet length will exclude both Layer 1 and 2 headers.

For example, if you want to send a packet length Layer 1 of 128 bytes, then the OSN will receive only 128 bytes packet.

If you want to send a Layer 2 packet length of 128 bytes, the OSN device will receive 128+20= 148 bytes packet, then the throughput can`t reach 100%, because an OSN device we have the Layer 1 header which occupies bandwidth.

Regarding the MPLS encapsulation: on the OSN device we configured the VPLS service, then when the packet reaches OSN device, the MPLS header will be added to the packet. The MPLS header has 26 bytes like below:

The whole original packet will be put in the payload field in the MPLS packet, then the header will occupy another bandwidth.

Now we calculate the throughput on a OSN device, with the next formula:

Packet length/real packet length in OSN = Throughput / Physical bandwidth.

When a tester is used on the Layer1 traffic, the packet length of 128 bytes will be:

Throughput = (128byte/(128byte+MPLS header))*physical bandwidth = 128/(128+26)*1Gbps = 831.2Mbps

When the tester is used on the Layer 2 traffic, the packet length of 128 bytes will be:

Throughput = (128bytes/(128bytes + MPLS header + Layer 1 header))*physical bandwidth = 128/(128+26+20)*1Gbps = 735.6Mbps

Conclusion: the packet send by the end user has some other encapsulation on our OSN device, then the throughput for the end user can`t be 100%. The longer length end user has, the bigger throughput he will have, as the encapsulation is a fixed length.

[hongtelecom-blog]

Optix OSN1800 Version description of the LSX board

OptiX OSN 1800 integrates both OTN and WDM features and uniformly transmits various services at the metro network edge, such as broadband, leased line, and mobile services. OptiX OSN 1800 supports unified transmission of multiple services ranging from 100 Mbit/s to 10 Gbit/s, unified switching of OTN/PKT/SDH services, and automatic commissioning at the optical layer. In this way, OptiX OSN 1800…

View On WordPress

[classyfoxwombat-blog-blog]

Telecomate supply Huawei OptiX OSN Series products and modules, select Huawei OSN 8800, OSN 6800, OSN 1800 for quote and compare specifications

[thunderlinkcom]

Attention: Abnormal OSC Communication of the OptiX OSN 1800

The networking is as follows: Station A and station B are configured with the LWX2 and the DMD2S. Each SCC is configured with one 1551-nm optical module to implement the OSC communication. The configuration of the two stations is the same. The NE version is 5.67.1.22. After Huawei optical power commissioning is complete, the ECC between station A and station B fails.  1.       The optical power is incorrect. 2.       The optical module is faulty. 3.       The SCC software is defective The SCC software has a bug. When the PHY chip used by the OSC optical port on the SCC works in BASE-EFX mode, special requirements are raised for the optical module. That is, the chip can be linked only when the amplitude of the electrical input signals of the optical module is higher than 800 mV. The amplitude of the electrical input signals of the currently used optical module is about 800 mV. Therefore, the OSC communication may fail.

The LWX2 does not support the ESC. Station A and station B use the optical module on the SCC to implement the OSC communication. Check the receive optical power of the RM1/TM1 optical port on the SCC. The check result shows that the receive optical power is normal. There is no abnormal alarm. Run :cm-get-eccroute. The remote NE cannot be queried. Run :cm-get-bdinfo to query the DCC state. The query result shows that the DCC state is rx-f. Reset the SCC. The fault persists.

Install OptiX OSN 1800 A and OptiX OSN 1800 B in the same equipment room. Interconnect the RM1s/TM1s of the two OptiX OSN 1800s with the added fixed optical attenuators through fibers. Run :cm-get-eccroute. The remote NE still cannot be queried. The DCC state is still rx-f.

The confirmation result by the related experts shows that the SCC software has a bug. When the PHY chip used by the OSC optical port on the SCC works in BASE-EFX mode, special requirements are raised for the optical module. That is, the chip can be linked only when the amplitude of the electrical input signals of the optical module is higher than 800 mV. The amplitude of the electrical input signals of the currently used optical module is about 800 mV. Therefore, the OSC communication may fail.

Software mitigation is performed in the latest version (5.67.01.24). The problem is solved. After the NE software is upgraded to this version, the OSC communication is normal on Huawei WDMOSN 1800.