Errata

JUNOS Enterprise Switching

Errata for JUNOS Enterprise Switching

Submit your own errata for this product.

The errata list is a list of errors and their corrections that were found after the product was released. If the error was corrected in a later version or reprint the date of the correction will be displayed in the column titled "Date Corrected".

The following errata were submitted by our customers and approved as valid errors by the author or editor.

Color key: Serious technical mistake Minor technical mistake Language or formatting error Typo Question Note Update

Version Location Description Submitted By Date submitted Date corrected
Printed
Page 148
4th paragraph (directly before the Auto-Negotiation discussion box)

4th sentence reads: The non-blocking nature of EX switches means they have no need to actually generate flow control, so this option simply allows the EX to reach to pause frames sent by a congested switch.

I think the last bit should read:

...generate flow control, so this option simply allows the EX to REACT to pause frames sent by a congested NIC.

Note from the Author or Editor:
Change:

"The non-blocking nature of EX switches means they have no need to actually generate flow control, so this option simply allows the EX to reach to pause frames sent by a congested switch.
"

to:

"The non-blocking nature of EX switches means they have no need to actually generate flow control, so this option simply allows the EX to react to pause frames sent by a congested switch.
"

JT  Oct 28, 2009 
Printed
Page 369
Last paragraph

First sentence of last paragraph:
"MSTP allows VLANs 501-1000 to be classified into one instance, and VLANs 1-500 to be classified into another region..."

I believe this was meant to read:
"and VLANs 1-500 to be classified into another INSTANCE..."

As instance and region are two distinctly different components of this topic, this can be confusing.

Paul Zugnoni  Oct 14, 2009 
Printed
Page 480
first section, diagram and next few pages

Lists reject as a valid terminating action for a firewall filter and reject is shown in the diagram and listed in the next few pages. Unlike the routing platforms the EX doesn't support a terminating action of reject in a firewall filter.

Note from the Author or Editor:
Remove reject from the list on page 480

Jason Schwegel  Oct 12, 2009 
Printed
Page 565
2nd section

sentence reads...You can extend the total number of PoE ports for an EX switch by inserting additional PoE cards if the power supply supports it.




Note from the Author or Editor:
now reads:
"You can extend the total number of PoE ports for an EX switch by inserting additional PoE cards if the power supply supports it"

Change to:

"You cannot extend the number of avaialble PoE ports beyond what the base model supports, but you can chhose not to use some PoE ports, or install a larger power supply, which can influance how much power is made avaialble to a given PoE port."

Jason Schwegel  Oct 08, 2009 
Printed
Page 553
3rd section

Interface ge-0/0/0 is set to support MAC-based RADIUS...but should be ge-0/0/9

Note from the Author or Editor:
Now reads:

"Interface ge-0/0/0 is set to support ..."

Change to:

"Interface ge-0/0/9 is set to support ..."

Jason Schwegel  Oct 08, 2009 
Printed
Page 48
2nd paragraph

The material is correct for the TX matrix consisting of 4 T640s or the 4 T1600s, both of which provide 2.5Tbps as written. However, the newer TX+ Matrix (9.6) provides up to 25Tbps. You may want to mention this.

Note from the Author or Editor:
Now reads:

"...or clustered as part of a TX matrix to scale up to 2.5 Tbps!"

Change to:

"or clustered as part of a TX matrix to scale up to 2.5 Tbps! If that'ts somehow not enough, the newest TX matrix offering, the TX matrix+ scales to 25Tbps.

Anonymous  Oct 05, 2009 
Printed
Page 499
Last 3 sentences on page

Book states that 100Mbps divided by 8 is 12.5 bytes per second. This should be 12.5 megabytes per second. Book then has user multiply by 1000 to convert bytes per second to bytes per millisecond. Should read divide by 1000 to convert bytes/sec to bytes/millisec. Interestingly, end result is the same, 12,500 bytes per millisecond.

Note from the Author or Editor:
"The result is 12.5 MB per second. Then divide it by 1,000 to get bytes per millisecond..."

Anonymous  Sep 07, 2009 
Printed
Page 271
5th paragraph

This paragraph referrs to ATM Virtual Circuit's as "Virtual Chassis"

Note from the Author or Editor:
This is confirmed. The correct expansion for "VC"s in this context is "Virtual Circuits".

LapTop006  Sep 06, 2009 
Printed
Page 192
Notice above 4th paragraphj

Although Juniper may consider the cables proprietry they are standard PCI-Express x8 external cables.

A clarification similar to that of SFP's that Juniper will not support non-Juniper cables would be better.

Note from the Author or Editor:
now reads:

"The VCP cables use a 68-pin connector and are considered proprietary and therefore only available through Juniper Networks and its authorized resellers. The user manual provides pin outs for the VCP cables, however."

Change to:

"The VCP cables use a 68-pin connector that are based on PCI-Express x8 technology. Whil it may be possible to obtain VCP cables from an external soure, Juniper considers them proprietary and does not support VCP applications based on such 3rd party cables. The user manual provides pin outs for the VCP cables, however."

LapTop006  Sep 06, 2009 
Printed
Page 14
Sidebar - Opportunistic Versus Deterministic MAC

"The widespread use of FD Ethernet, enabled by switches, has eliminated the potential for collisions," and the 1024 host limit.

Note from the Author or Editor:
now reads:
"The wide spread use of FD ethernet, enabled by switches, has eliminated the potential for collisions, which simply made a good thing all the better. "

Change to:

"The wide spread use of FD ethernet, enabled by switches, has eliminated the potential for collisions (and with it the need for a 1,024 node limit), which simply made a good thing all the better. The relaxation of teh node limit is rather moot, however, given that modern Etherent is now switched rather than repeated, and that switches terminate colision domains, the modern reality is that Ethernet has become a point-to-point technology that really only nees to support two nodes per link anyway.

LapTop006  Sep 06, 2009 
Printed
Page 68
1st paragraph under "Layer 2 switching" sub topic in last sentence

"... given logical unit is configured with family bridging ..." where "bridging" should be "ethernet-switching". Unless an older version of code used this stanza, but according to JUNOS 9.5 "bridging" is not an option.

Note from the Author or Editor:
This is confirmed. "family bridging" should be changed to "family ethernet-switching".

Shane Dawalt  Aug 28, 2009 
Printed
Page 65
3rd paragraph under "EX3200 Architecture" in last sentence

"The CPU also drives the 1 Gbps em0 ..." The OoB interface name is me0.

Note from the Author or Editor:
Confirmed. Change "em0" to "me0", 2nd to last paragraph

Shane Dawalt  Aug 28, 2009 
Printed
Page 54
Figure 2.4

In the text describing Figure 2.4, references are made to "distribution layer", but the figure has the distribution layer named "aggregation" which is somewhat confusing.

Note from the Author or Editor:
Confirmed. The text should change to match the figure, or the figure 2-4 can be modified to have "Aggregation/Distribution", which is what i would prefer.

Shane Dawalt  Aug 28, 2009 
Printed
Page 149
2nd paragraph

Last sentence within Auto-Negotiation Stance section says:

"This is due to lack of auto-negotiation, and it causes the switch to fall back to the lowest ability of HD..."

Since the switch is fixed at 100 FD, the switch cannot fall back; it must be the network card that is in auto-negotiation state that falls back to HD.

Note from the Author or Editor:
I have reviewed an agree we need to change to:

"This is due to lack of auto-negotiation, and it causes the NIC to fall back it's lowest capability, which ...""

Shane Dawalt  Aug 26, 2009 
Printed
Page 51
2nd section

It says two 54Gbps when it should read two 64Gbps Virtual Chassis Port (VCP) ports

Note from the Author or Editor:
Change "54" to "64", 1st para, 5th line.

Anonymous  Aug 24, 2009