Too Many HTML Connections Message

[waynes]

When trying to access a controller through System Manager, I am getting the message "Too many HTML connections".  Has anybody else seen this?  Is there a way to determine who or what connections are connected to the filer through http or https?

[chris_hague]

Anther response from netapp regarding this issue... Appears to be a staff only bug...

"I have confirmed bug#552607 to be the issue here. I do not have a clear timeline for the fix but Engineering is working on this.

You can sign up for bug Watch here http://support.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=552607 to receive notifications."

Anyone else had a reply?

[lee_harrison]

NetApp have told me that using the 'httpstat' utility is not the way to monitor this issue and went on to say the only way to resolve was to failover Controllers

[chris_hague]

Useful when it happens multiple times a day if you use OCSM or SMVI!

[rmharwood]

We're also being affected by this - FAS3160 running 8.1RC3 and DFM 5.0. In 24 hours the httpd max has hit the limit of 1023. Any updates since last week on this?

[thommy_lan]

Are you running SMVI?

We had this problem but after a uninstall of SMVI the problem was gone.

[rmharwood]

Actually yes. This is the main reason I switched to DFM 5 / OnCommand Core/Host in the first place. I'll see if I can go about removing it.

I feel very disappointed with a lot of the recent software changes from NetApp.

[MARTINLEGGATT]

Hi,

We are using DFM 5.0 and Oncommand but not SMVI.  The httpstat counter did drop back to normal for a few days but has since begun hitting 1023 again and the performance monitoring in DFM is intermittent as before.

Regards

Martin

[danpancamo]

Ran into this problem too...   I don't see a fix yet...

NetApp Release 8.0.2 7-Mode

[keeshuizer]

Is there any news about this problem, we seem to be having this problem since a couple of days, rebooted our filer yesterday which solved the problem for a while, but the counter is now already back at 1023, and logging in to the filer using the system manager hangs on the login screen

[chris_hague]

Apparently this is fixed in 8.1 (and possibly 8.0.3) however after trying 8.1 in a test environment I have seen the same issue occuring

Spent 8 hours trying to get one of our live systems (8.0.2) to allow connections from either System Manager or the VSC by killing everything HTTPS which was connecting to the filer and then waiting for the stats to drop from 1023 to a sensible level, only for the issue to occur again within minutes of it starting to work again.

We'll upgrade the live system to 8.1 if advised by netapp to do so.

[thommy_lan]

Can comfirm  that this issue still exists on DataOnTap 8.1.

We running a FAS3140 with 8.1 and DFM 5.0 and when we hit 1023 connections our snapvault jobs via Protection Manager fails.

How many of you that have this problem have made a case at Netapp?

We all need to do that so Netapp can take this issue seriuosly.

[joekr1234]

We're experiencing the exact same issue on only one of four filers.  NetApp, do you you have a fix for this?

[KEITH_TRAN]

We've been experiencing the same exact issue, but in addendum we've had other variance manifest. Have others with this issue had other issues besides GUI management access?

[jurgen_wellens]

We had the same issue after a multi-disk panic this morning.

Were able to solve the issue by performing the following command on dfm server: "dfm ssl service reload".

[joekr1234]

This is now impacting another filer..  During our weekly meeting with NetApp today, I was informed this issue has been escalated to engineering..

[richard_may]

8.1 on a 2040.  Just experienced this issue for the first time.  No fix found yet.

NetApp!  

[benhallnow]

Hey folks,

I am working with one of the customers that hit BURT 552607. I just got an update from NetApp Support that the fix will be in ONTAP 8.1.2, and they are also pushing for an 8.1.1 P-release to include the fix.

Thanks,

Ben

[joekr1234]

Ben,

Appreciate the update.  Are you a NetApp employee?

While the fix will be welcome, I want a specific technical reason for this issue.  I've got approx 2.6 Pb behind V-series and another couple 100 Tbs of NAS filers.  An upgrade to Ontap version X isn't something I can just decide to push out in a day or two...  These are client facing filers and maintenance windows are few and far between.

Thanks,

Joe

[benhallnow]

Joe,

I am a NetApp partner, and have been working with my NetApp counterparts to get this issue escalated on my customer's behalf. The reason it is so critical in my customer's case is that when this problem happens, SMVI (AKA VSC) snapshot jobs fail, because it too makes calls to the NetApp via httpd.

The basic explanation I got from Support is that a legacy API call is being made to httpd, which does not properly handle/clear connections. The fix will allow the connections to timeout/clear out properly without hitting the max connection threshold.

Thanks,

Ben

[joekr1234]

Ben,

Great information, terrible situation for you client.  With 1668 views of this posting, I would hope that *someone* from NetApp would be giving us some regular updates...

Thanks,

Joe

[joekr1234]

Public BURT is out...

http://support.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=552607

I've been told it's going to be over a month before code is ready for deployment....