Enabled HTTP acceleration - Lots of error messages.

[RRV]

Wonder if anybody can please share some light on what these error messages actually mean:

E13 L1L2 DCID:0 - re-inject error: size=48, dest=XXX.XXX.XXX.XXX, err=S_errno_EHOSTDOWN, g_dcSockRaw = 23

E05 Error: can only write word-aligned memory

I39 HTTP - First element write error - ret=8 key=5316ce749ed6488e11bb36016855c13c uri=s/img/beyondandes315.jpg len=4/0 trans=0x5719cbac

E39 closeIoHandle(trans=0x5719cbac) - unable to delete object (key www.abc.net.auhttp://www.abc.net.au/tv/programs/img/beyondandes315.jpg) - ret=15 where=281

E39 closeIoHandle(trans=0x5719cbac) - Error found (err=8) - deleting object key abc.net.auhttp://www.abc.net.au/tv/programs/img/beyondandes315.jpg (281)

E50 write-6696: OC_OUT_OF_RESOURCE

Thanks.

Rafael 

Jcare case: Case #2009-0505-0637 - HTTP acceleration causes loss of HTTP traffic

[Antares]

Hey RRV,

I see that you raised a JTAC case already and an engineer from Advanced TAC

already provided some explanations.

I39 HTTP - First element write error - ret=8 key=5316ce749ed6488e11bb36016855c13c uri=s/img/beyondandes315.jpg len=4/0 trans=0x5719cbac

>> The above message indicates that the HTTP module is not able to write

the HTTP response into the OS cache. The possible reason behind getting

this error message is that the OS cache is full and all objects are

being actively accessed so LRU can not happen. The implication of this

message is that the flow will be soft quit. Acceleration will not be

observed for the object.

E50 write-6696: OC_OUT_OF_RESOURCE

>>  The above message is seen when the object store has run out of cache

RAM. It could be an expected situation because of heavy cache usage or

an error.

So I would strongly suggest you follow up directly with the engineer by updating

the case should you have further questions.

Thanks.

[RRV]

Good Evening Juniper Employee.

Thanks for reposting JCare notes to this forum.

 Any comments regarding the rest of listed errors?

About your recommendation : " ... strongly suggest you follow up directly with the engineer by updating

the case should you have further questions ..."

 afraid it feels like Jcare engineer forwarded it up to Developers and kinda given up on it himself.

So I was curious if anybody else had this problem and can share his/her experience.

Cheers

Rafael

[Antares]

Hello RRV,

Regarding the rest:

E13 L1L2 DCID:0 - re-inject error: size=48, dest=XXX.XXX.XXX.XXX, err=S_errno_EHOSTDOWN, g_dcSockRaw = 23

>> This has nothing to do with HTTP (as you may probably know already) but that just means the WX is unable

to decompress and deliver traffic to host XXX.XXX.XXX.XXX. Probably the host is down, or it's a user with a laptop who left

or a remote application scanning a range of IP addresses etc...

E39 closeIoHandle(trans=0x5719cbac) - unable to delete object (key www.abc.net.auhttp://www.abc.net.au/tv/programs/img/beyondandes315.jpg) - ret=15 where=281

E39 closeIoHandle(trans=0x5719cbac) - Error found (err=8) - deleting object key abc.net.auhttp://www.abc.net.au/tv/programs/img/beyondandes315.jpg (281)

>> HTTP module tries to close cache object file but there is an error in the close operation. The file is closed when OS I/O operation is successful or if any error has occured (which is your case here)

Obviously if this is causing grief and it's a bug, then we need to get it fixed

So again, please follow up with the engineer.

Thanks. 

[RRV]

Hi Antares.

Thanx

Rafael