Win13??

Any chance of this server getting fixed sometime soon? It seems almost daily a pool needs to be reset or the server restarted. Not the best way to handle an issue and it is becoming frustrating as our sites go down for 10-15 mins at a time.

Shooner,

IIS was reset for the first time in over 50 hours, this just happens sometimes.

Pools recycle a few times a day, it is the way IIS6 is designed.

It is not daily that iisreset it is run, it just typically happens when I am in, I will talk to the earlier shift about doing it around 7am EDT instead.

I’m not complainign about pool resets, its the fact that my sites go down, I contact live chat and they end up resettign them or the server to fix the issue. Something is causing issues with pools on that server and has it seems since day 1. We keep getting outage reports almost daily that our sites are down. Typically they are back up in 15 mins or so if we just let things happen on their own, but our clients are not always pleased when sites are down mid day. Other servers don’t seem to have these issues or at least as often.

PM or send a ticket with site names and I will move to your own pool, that way they can be isolated from anyone else. I moved one of yours once, but it is still in a shared pool.

Thanks Stepehen. I’m just heading out but will send you a list later today or this weekend.

Some of our domains on win13 were working while others were down so my assumption was a pool issue.

The server also seemed MUCH faster after the reboot.

not reboot, just iisreset :slight_smile:

Just learned a little bit more, I think one of the sites on win13 banned a user from forums, and then they launched some proxy based GET attack causing the issues today.

It was odd when I did iisreset, as cpu was sitting at 11% and ram at 54% usage, all actually below normal for this time of day. IT was just taking a beating from so many requests.

This has stopped for now, and I will be closely monitoring for anything like it and blocking all IPs involved in such an attack again at the router.

Thanks Stephen. I suspected something had to be up with it. Hopefully you have it solved now. I’ll PM you my domains as well if you want to move them into a seperate pool

Not sure if you guys are working on something on Win13 but right now we cannot access any sites on that server. Other sites on other Jodohost servers are fine. :frowning:

Apparently was a pool recycle. They seem to be alive again now.

I have got to discuss this with the techs, something is happening every time I am off to make all pools recycle at the SAME time, this is very bad, as it causes the recycle to take 2-8min per pool, a typical pool recycle really is not noticed, only when they all go down at once getting recycled is it a problem.

I know this is happening because I can see in management tools that all the processes are 3 hours 2 min old right now.

This never happens while I am on shift(all pools at once), so I need to find what they are doing to cause this to happen.

Sites down yet again this morning. Some work, some are slow, and other not loading at all now for 5 mins.

As we are in Atlantic time we need sites live by at least 5AM EST. As is we are finding sites down in the mornig nhere. Not sure it is the pool issue or what is going on but it is becoming all too common on Win13.

And now it would seem all our WIN13 sites are down. ?( and no one available on live chat. Frustrating.
Sites not on Win13 always seem to be fine at same time.

No we are working on it. Sites are up right now but slow
Prakash is working to isolate the cause of a sudden spike in CPU. Most likely another abusing script :frowning:

we should have the speed normal very shortly

LiveChat is available.. however yes its busy right now.

WEll they are up now but speeds are inconsistent. Seems the pool was hung yet again which seems to be a daily event now on Win13. This server needs to be fixed, we have had nothign but issues with it since day 1. It is all too common for sites to be down or slow on it. Very frustrating.

I am monitoring this for many days now. We mailed the user who was responsible for the site abusing the server, but havent yet had a response back. His site
was taking 50% of CPU & causing server to hang most of the times. I have disabled that domain now.

Prakash Sharma

schooner, I assure you we are committed to ensuring Win13 performs at the same level of our other servers. I understand some of these abuse issues are frustrating

We are I guess a little lenient with repeat abusers. I’ll have a discussion with Stephen in the morning about our policy in this regard

Ok thanks. This is a daily occurence now for us. Between the pools recycling at key times for us and the abuse it is becoming a headache for us on this server. It is al most a sure thign that our sites are down when we check them in the morning :frowning: Then 30 mins later after live chat and forum posts they are fixed; we jsut can’t continue like this. We need to have confidence the server jsut is working.

I am checking win13, and can see that prakash in fact did not recycle the pools this morning, all of them have been up for a numb er of hours now, this is good.

However I have CC’s of some mails he sent to customers about their access db causing a look and using continual 50-80% CPU, the csutoemr has replied and the site is being changed to MSSQL, in the meantime the files causing the loop have all been renamed.

This is some maintenance opteration this site runs, always about 4-7am EDT, it casued major issues another time, and we asked them to move to mssql ASAP, this time it is still on access and we have asked not to rename the abusing files until it is on MSSQL.

Access just is not made for a serious web app.

I am very sorry about this, but glad to see the POOLS in fact were fine today, and have been the last 2 days, just the abuse issue today was very bad, not allowing other pools to even run for the most part.

Ah thansk Stephen. I just assumed it was hte pool as it was acting like that with some sites fine and other not and it was in the same timeframe. Hopfully now the issue is solved.