Hmm, I was wondering, if Tracerts and Pings, the ms reading is higher, .... will this mean its not good for Singapore?
From my Tracert for both Hawkhost and Stablehost, Code:
TraceRoute to 50.23.66.3 [sjc.hawkhost.com] Code:
TraceRoute to 216.12.197.183 [sng.hawkhost.com] Code:
TraceRoute to 199.96.156.70 [cp22.stablehost.com] Code:
Ping 199.96.156.70 [cp22.stablehost.com] Code:
Ping 50.23.66.3 [sjc.hawkhost.com] Code:
Ping 216.12.197.183 [sng.hawkhost.com] Now I'm confused... haha Ping and Tracert -- Stablehost is the winner... but what is the significance of these numbers? |
Understand that pings and traceroutes are not measurements of speed.
A ping is simply a question. Your computer asks "Server, are you there?" and the server either responds "Hello, here I am!" or there's no response. No response can mean (A) downtime, or (B) ICMP packets are blocked. ICMP packets are what a "ping" is using to communicate with. It's a verification of communication, not a speed test. Traceroute is a method to see how communications is routing from router/switch/datacenter to router/switch/datacenter, along the path between your current location and the server being requested. It's drawing a map for you, showing how to navigate the maze to get from point A to point B. It's not a measure of speed. Some points along the path may block ICMP packets, so tracerts commonly have timeouts between the start and end. At most, ping and tracert can be used to analyze latency, which is speed of requests back and forth. Unless you're running a game server, a streaming videos archive, or some sort of stock exchange application, latency isn't really an issue. For websites, it's really not that important, as long as you're within acceptable norms. We are, after all, talking in milliseconds -- 1/1000th of a second. So 130ms = 0.13 seconds. (Also understand browsers load content concurrently, so you cannot simply multiply latency times the number of requests to arrive at site speed. The math is not that simplistic!) Speed is effected by routing times, but the ultimate determining factor of speed is the content itself. This server is 130ms from my location, but pages load in under 1 second. The forum loads in under 1.5 seconds. Both of those are excellent speeds. I can ping my ISPs server in 35ms, but its homepage loads like crap, and takes 2-3 seconds (at best). This is why server-side caching, and leveraging user-end caching (browser cache), is so important. Anything above 150ms should be a concern. Anything in the 90-150ms range is on the higher end of acceptable, and would quality as "good". Anything below 90ms is "excellent". Opinions on what makes a "good" or "bad" number varies. And understand some people are clueless about what any of it means, and just give out random numbers, so be careful who you listen to. (There's a lot of inexperienced teens and college students out there writing on blogs and forums, pretending to be experts.) You also need to test from your current physical location -- not using some online tool. Bring up DOS and run tracert against a domain (FQDN) or IP address. Same for ping. See what your local results are. Then test against from a multiple-location tool, such as Just-Ping.com. The Hawkhost Singapore location is both slow, and losing lots of packets. They're in the Softlayer Singapore DC, and the Softlayer SG test server is fine. So I don't know what's wrong with the Hawkhost SG server at this moment. |
Temp post to be moved
Tracert of Stablehost and Hawkhost -- vs -- our JaguarPC VPS node:
Code:
TraceRoute to 199.96.156.70 [cp22.stablehost.com] Code:
TraceRoute to 216.12.197.183 [sng.hawkhost.com] - - Hawkhost Singapore Code:
TraceRoute to 50.23.66.3 [sjc.hawkhost.com] - Hawkhost CA San Jose Code:
TraceRoute to 209.140.18.169 [cybermind.com.sg] -- JPC Pings Stablehost / Hawkhost VS JPC's our own node: Code:
Ping 199.96.156.70 [cp22.stablehost.com] Code:
Ping 216.12.197.183 [sng.hawkhost.com] Code:
Ping 50.23.66.3 [sjc.hawkhost.com] Code:
Ping 209.140.18.169 [cybermind.com.sg] |
The pings and traceroutes between JaguarPC and Stablehost / Hawkhost all look normal, to be honest.
And cybermind.com.sg loads extremely fast in a web browser. |
ok, here goes! hot from my PC
Code:
Tracing route to sjc.hawkhost.com [50.23.66.3] Code:
Tracing route to sng.hawkhost.com [216.12.197.183] Code:
Tracing route to cp22.stablehost.com [199.96.156.70] Code:
Tracing route to cybermind.com.sg [209.140.18.169] |
Pings hot from my PC! * Sweating by the "oven" *
Code:
Pinging sjc.hawkhost.com [50.23.66.3] with 32 bytes of data: Code:
Pinging sng.hawkhost.com [216.12.197.183] with 32 bytes of data: Code:
Pinging cp22.stablehost.com [199.96.156.70] with 32 bytes of data: Code:
Pinging cybermind.com.sg [209.140.18.169] with 32 bytes of data: |
There's nothing unusual about the number of hops on the JaguarPC route. Start to worry only when it exceeds 30, or if it seems to loop back to itself for no apparent reason.
If anything, the Stablehost speeds look oddly slower than I would have expected. But then again, that can be due to any number of network related issues -- at your end, at their end, or in between. The Hawkhost SG server does appear a bit slow (on the tracert), given the size of Singapore, and the fact that you're almost on top of it. The networks there must be slower or more congested than I thought. It did ping faster, but it usually does. Run UptimeScout against your site, and see how the site homepage actually loads starting on the second load. I have a feeling it's going to be faster than you think. |
From the Multi-location, I'll just insert the SG numbers -- here for simplicity :
SJC - Hawkhost : Singapore, Singapore: Packets lost (80%) 185.9 188.2 190.5 50.23.66.3 SNG - Hawkhost : Singapore, Singapore: Packets lost (60%) 34.6 35.7 36.4 216.12.197.183 CP22 -Stablehost : Singapore, Singapore: Okay 217.2 231.8 243.8 199.96.156.70 JPC -- own : Singapore, Singapore: Okay 281.4 282.3 285.4 209.140.18.169 Seems the SG DC by Hawkhost wins again! |
I think the weak link with the Los Angeles JaguarPC location is they're using Cogent bandwidth -- or rather the datacenter is (Colo@ in LA). Cogent has long been known as a discount bandwidth supplier, and sometimes their routes are screwy. Increased latency is par for the course with Cogent. It doesn't make a huge difference domestically, for websites, but it may be a factor in access quality from overseas.
Hawkhost is going over nLayer. Stablehost is going over Highwinds to CWIE. |
Wah.... lol... . should have posted all this information ... (researched), before getting JPC for him! Damn!
So now the case is Hawkhost or Stablehost? The Names you posted means nothing to me.. haha Hawkhost SG -- according to the tracerts and pings, seems ok (I'm testing this over a 3G connection -- usb stick).... But if the Story about Tokyo routing is true.. it should be seen here as well? I saw in their blog -- announcement for The SG DC, someone posted the tracert results there, and NTT Tokyo came up.. I saw a familiar one in Stablehost Tracert: 4 100 ms 79 ms 97 ms anutsi11.starhub.net.sg [203.118.3.227] That's our Singapore Starhub ISP! |
Installed Uptime Scout .. here are my findings :
Cybermind - JPC : 3777ms SJC - Hawkhost : 725ms SNG - Hawkhost : 347ms Stablehost : 875ms |
Even I don't always think of bandwidth carriers when looking for new hosts. But I've gotten better about it in recent years.
Hawkhost and Stablehost are fine. Had their carriers been an issue, I'd have mentioned it. Understand that I'm not saying Cogent is "bad" as much as it's just not the best. And in their specific case, the routes and latency is a known issue for most of the past decade. Folks often thought Cogent would grow out of it, as the network matured, but it never happened. You get what you pay for. JaguarPC in Atlanta -- their main datacenter -- runs over nLayer. Your UptimeScout JPC reading sucks. 4 seconds... yuck. But is that the first load, or the second and beyond? |
That's the first load.. over the last 5 loads (I set to auto run every 10 min) -- range between 3019 to 4383 ms!!!!
And AS I can see, SNG Hawkhost is clearly the winner here : Ranging from 207 to 347ms! Stable host : 596 to 875ms.. In this regard, SJC Hawkhost is slightly better than Stablehost only... so the message is clear? -- SNG Hawkhost -- SG!! ** Hmmm, we used to be from Atlanta DC previously, but that sucks as well... Just curious and run the Uptime Scout with my own shared hosting with Arvixe, and that was even worst... 4752ms! |
Arvixe is a good-but-not-best host. On our soon-to-be-released 1st/2nd/3rd/4th class scale, they're definite 2nd Class host. In my opinion, the biggest problem is that Arvand (the owner) can be a real ass (unprofessionally rude) to customers on public forums, and he needs to leave customer relations to others within the organization. Outside of that their techs and servers are quite decent. Prices, too.
I think Hawkhost @Singapore is a good option for your needs. :) |
OK.
I'm shifting my Arvixe hosting to JPC for the time being till our contract with JPC ends... These few days, Arvixe keep having internal 500 errors!! Up n down several times... As for Hawkhost... if they have recurring discount for shared hosting.... now they don't... Also their new Semi dedicated hosting is only available at Texas - Dalias.... So I guess that is out for my friend... Stablehost in Arizona and Hawkhost in Texas had similar response times.... Hmm.... pondering.. |
Well, again, remember that ping/traceroute response times are not necessarily measurements of "speed". ;)
Regarding coupons: - Current Hawkhost 50% OFF coupon whts50v2 does not appear to be limited to any one location, and can be use for any payment term. - Current Stablehost 25% OFF coupon US-WHT2012-25 is good for USA shared plans, on any payment plan. |
just out of curiousity... I tried entering whts50v2 in hawkhost promotion code box but it doesn't apply for semi dedicated in Texas...
We have till next year Sep to change host.... though.... You're saying pings and tracert doesn't indicate the speed and reliability... so what else will help us determine the good ones from the bad? |
The content of the site ultimately determines the speed -- latency is an effect, but not the ulterior determining factor. There's too much weight put into latency (and thus ping and traceroute), instead of ascertaining the entire "big picture" of what's going on. And that big picture includes the coding quality of the site, any applicable caching plugins/modules, and things of that nature.
Some hosts overstuff (overload) servers -- NOT to be confused with "overselling" (which is a poor way to express reasonable resource/server management) -- and that can limit the way code functions on a certain site, at a specific host. The only way to truly determine good hosts from bad is to use them, and see how it works. Of course, the ability to judge a host's quality is limited by your own understanding of networking, servers, and site coding. That's why non-affiliate-biased lists of "good hosts" (like the Best Hosts List here at digitalFAQ.com) is so important to folks like yourself, as they've been created by a person/persons with quite a bit of experience and knowledge in these many variables, including how they interplay with one another. In terms of deciding between Stablehost and Hawkhost -- both are good, so nothing to fear. Neither overload servers. In terms of deciding the best location, latency can assist in that. It does appear that Hawkhost Singapore would be ideal for your Singapore traffic. The Hawkhost semi-dedicated coupon is sdlaunch and is good for 30% for life on all semi-dedicated plans. :) |
OK.
hmm too bad their semi-dedicated are only available in Texas.... Thanks for your guidance! |
Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.